Compare commits

...

16 Commits

Author SHA1 Message Date
Arseny Sher
910640c955 Update sk_collect_dumps to new inventories, auth and LSN formatting. 2023-12-19 17:16:01 +03:00
Arseny Sher
b701394d7a Fix WAL waiting in walproposer for v16.
Just preparing cv right before waiting is not enough as we might have already
missed the flushptr change & wakeup, so re-checked before sleep.

https://neondb.slack.com/archives/C03QLRH7PPD/p1702830965396619?thread_ts=1702756761.836649&cid=C03QLRH7PPD
2023-12-19 15:34:14 +04:00
John Spray
d89af4cf8e pageserver: downgrade 'connection reset' WAL errors (#6181)
This squashes a particularly noisy warn-level log that occurs when
safekeepers are restarted.

Unfortunately the error type from `tonic` doesn't provide a neat way of
matching this, so we use a string comparison
2023-12-19 10:38:00 +00:00
Christian Schwarz
6ffbbb2e02 include timeline ids in tenant details response (#6166)
Part of getpage@lsn benchmark epic:
https://github.com/neondatabase/neon/issues/5771

This allows getting the list of tenants and timelines without triggering
initial logical size calculation by requesting the timeline details API
response, which would skew our results.
2023-12-19 10:32:51 +00:00
Arpad Müller
fbb979d5e3 remote_storage: move shared utilities for S3 and Azure into common module (#6176)
The PR does two things:

* move the util functions present in the remote_storage Azure and S3
test files into a shared one, deduplicating them.
* add a `s3_upload_download_works` test as a copy of the Azure test

The goal is mainly to fight duplication and make the code a little bit
more generic (like removing mentions of s3 and azure from function
names).

This is a first step towards #6146.
2023-12-19 11:29:50 +01:00
Arpad Müller
a89d6dc76e Always send a json response for timeline_get_lsn_by_timestamp (#6178)
As part of the transition laid out in
[this](https://github.com/neondatabase/cloud/pull/7553#discussion_r1370473911)
comment, don't read the `version` query parameter in
`timeline_get_lsn_by_timestamp`, but always return the structured json
response.

Follow-up of https://github.com/neondatabase/neon/pull/5608
2023-12-19 11:29:16 +01:00
Christian Schwarz
c272c68e5c RFC: Per-Tenant GetPage@LSN Throttling (#5648)
Implementation epic: https://github.com/neondatabase/neon/issues/5899
2023-12-19 11:20:56 +01:00
Anna Khanova
6e6e40dd7f Invalidate credentials on auth failure (#6171)
## Problem

If the user reset password, cache could receive this information only
after `ttl` minutes.

## Summary of changes

Invalidate password on auth failure.
2023-12-18 23:24:22 +01:00
Heikki Linnakangas
6939fc3db6 Remove declarations of non-existent global variables and functions
FileCacheMonitorMain was removed in commit b497d0094e.
2023-12-18 21:05:31 +02:00
Heikki Linnakangas
c4c48cfd63 Clean up #includes
- No need to include c.h, port.h or pg_config.h, they are included in
  postgres.h
- No need to include postgres.h in header files. Instead, the
  assumption in PostgreSQL is that all .c files include postgres.h.
- Reorder includes to alphabetical order, and system headers before
  pgsql headers
- Remove bunch of other unnecessary includes that got copy-pasted from
  one source file to another
2023-12-18 21:05:29 +02:00
Heikki Linnakangas
82215d20b0 Mark some variables 'static'
Move initialization of neon_redo_read_buffer_filter. This allows
marking it 'static', too.
2023-12-18 21:05:24 +02:00
Sasha Krassovsky
62737f3776 Grant BYPASSRLS and REPLICATION explicitly to neon_superuser roles 2023-12-18 10:54:14 -08:00
Christian Schwarz
1f9a7d1cd0 add a Rust client for Pageserver page_service (#6128)
Part of getpage@lsn benchmark epic:
https://github.com/neondatabase/neon/issues/5771

Stacked atop https://github.com/neondatabase/neon/pull/6145
2023-12-18 18:17:19 +00:00
John Spray
4ea4812ab2 tests: update python dependencies (#6164)
## Problem

Existing dependencies didn't work on Fedora 39 (python 3.12)

## Summary of changes

- Update pyyaml 6.0 -> 6.0.1
- Update yarl 1.8.2->1.9.4
- Update the `dnf install` line in README to include dependencies of
python packages (unrelated to upgrades, just noticed absences while
doing fresh pysync run)
2023-12-18 15:47:09 +00:00
Anna Khanova
00d90ce76a Added cache for get role secret (#6165)
## Problem

Currently if we are getting many consecutive connections to the same
user/ep we will send a lot of traffic to the console.

## Summary of changes

Cache with ttl=4min proxy_get_role_secret response.

Note: this is the temporary hack, notifier listener is WIP.
2023-12-18 16:04:47 +01:00
John Khvatov
33cb9a68f7 pageserver: Reduce tracing overhead in timeline::get (#6115)
## Problem

Compaction process (specifically the image layer reconstructions part)
is lagging behind wal ingest (at speed ~10-15MB/s) for medium-sized
tenants (30-50GB). CPU profile shows that significant amount of time
(see flamegraph) is being spent in `tracing::span::Span::new`.

mainline (commit: 0ba4cae491):

![reconstruct-mainline-0ba4cae491c2](https://github.com/neondatabase/neon/assets/289788/ebfd262e-5c97-4858-80c7-664a1dbcc59d)

## Summary of changes

By lowering the tracing level in get_value_reconstruct_data and
get_or_maybe_download from info to debug, we can reduce the overhead of
span creation in prod environments. On my system, this sped up the image
reconstruction process by 60% (from 14500 to 23160 page reconstruction
per sec)

pr:

![reconstruct-opt-2](https://github.com/neondatabase/neon/assets/289788/563a159b-8f2f-4300-b0a1-6cd66e7df769)


`create_image_layers()` (it's 1 CPU bound here) mainline vs pr:

![image](https://github.com/neondatabase/neon/assets/289788/a981e3cb-6df9-4882-8a94-95e99c35aa83)
2023-12-18 13:33:23 +00:00
47 changed files with 1229 additions and 866 deletions

9
Cargo.lock generated
View File

@@ -3145,6 +3145,7 @@ dependencies = [
"tokio",
"tokio-io-timeout",
"tokio-postgres",
"tokio-stream",
"tokio-tar",
"tokio-util",
"toml_edit",
@@ -3182,11 +3183,19 @@ dependencies = [
name = "pageserver_client"
version = "0.1.0"
dependencies = [
"anyhow",
"async-trait",
"bytes",
"futures",
"pageserver_api",
"postgres",
"reqwest",
"serde",
"thiserror",
"tokio",
"tokio-postgres",
"tokio-stream",
"tokio-util",
"utils",
"workspace_hack",
]

View File

@@ -29,13 +29,14 @@ See developer documentation in [SUMMARY.md](/docs/SUMMARY.md) for more informati
```bash
apt install build-essential libtool libreadline-dev zlib1g-dev flex bison libseccomp-dev \
libssl-dev clang pkg-config libpq-dev cmake postgresql-client protobuf-compiler \
libcurl4-openssl-dev openssl python-poetry lsof libicu-dev
libcurl4-openssl-dev openssl python3-poetry lsof libicu-dev
```
* On Fedora, these packages are needed:
```bash
dnf install flex bison readline-devel zlib-devel openssl-devel \
libseccomp-devel perl clang cmake postgresql postgresql-contrib protobuf-compiler \
protobuf-devel libcurl-devel openssl poetry lsof libicu-devel
protobuf-devel libcurl-devel openssl poetry lsof libicu-devel libpq-devel python3-devel \
libffi-devel
```
* On Arch based systems, these packages are needed:
```bash

View File

@@ -298,7 +298,7 @@ pub fn handle_roles(spec: &ComputeSpec, client: &mut Client) -> Result<()> {
// safe to add more permissions here. BYPASSRLS and REPLICATION are inherited
// from neon_superuser.
let mut query: String = format!(
"CREATE ROLE {} INHERIT CREATEROLE CREATEDB IN ROLE neon_superuser",
"CREATE ROLE {} INHERIT CREATEROLE CREATEDB BYPASSRLS REPLICATION IN ROLE neon_superuser",
name.pg_quote()
);
info!("role create query: '{}'", &query);

View File

@@ -0,0 +1,197 @@
# Per-Tenant GetPage@LSN Throttling
Author: Christian Schwarz
Date: Oct 24, 2023
## Summary
This RFC proposes per-tenant throttling of GetPage@LSN requests inside Pageserver
and the interactions with its client, i.e., the neon_smgr component in Compute.
The result of implementing & executing this RFC will be a fleet-wide upper limit for
**"the highest GetPage/second that Pageserver can support for a single tenant/shard"**.
## Background
### GetPage@LSN Request Flow
Pageserver exposes its `page_service.rs` as a libpq listener.
The Computes' `neon_smgr` module connects to that libpq listener.
Once a connection is established, the protocol allows Compute to request page images at a given LSN.
We call these requests GetPage@LSN requests, or GetPage requests for short.
Other request types can be sent, but these are low traffic compared to GetPage requests
and are not the concern of this RFC.
Pageserver associates one libpq connection with one tokio task.
Per connection/task, the pq protocol is handled by the common `postgres_backend` crate.
Its `run_message_loop` function invokes the `page_service` specific `impl<IO> postgres_backend::Handler<IO> for PageServerHandler`.
Requests are processed in the order in which they arrive via the TCP-based pq protocol.
So, there is no concurrent request processing within one connection/task.
There is a degree of natural pipelining:
Compute can "fill the pipe" by sending more than one GetPage request into the libpq TCP stream.
And Pageserver can fill the pipe with responses in the other direction.
Both directions are subject to the limit of tx/rx buffers, nodelay, TCP flow control, etc.
### GetPage@LSN Access Pattern
The Compute has its own hierarchy of caches, specifically `shared_buffers` and the `local file cache` (LFC).
Compute only issues GetPage requests to Pageserver if it encounters a miss in these caches.
If the working set stops fitting into Compute's caches, requests to Pageserver increase sharply -- the Compute starts *thrashing*.
## Motivation
In INC-69, a tenant issued 155k GetPage/second for a period of 10 minutes and 60k GetPage/second for a period of 3h,
then dropping to ca 18k GetPage/second for a period of 9h.
We noticed this because of an internal GetPage latency SLO burn rate alert, i.e.,
the request latency profile during this period significantly exceeded what was acceptable according to the internal SLO.
Sadly, we do not have the observability data to determine the impact of this tenant on other tenants on the same tenants.
However, here are some illustrative data points for the 155k period:
The tenant was responsible for >= 99% of the GetPage traffic and, frankly, the overall activity on this Pageserver instance.
We were serving pages at 10 Gb/s (`155k x 8 kbyte (PAGE_SZ) per second is 1.12GiB/s = 9.4Gb/s.`)
The CPU utilization of the instance was 75% user+system.
Pageserver page cache served 1.75M accesses/second at a hit rate of ca 90%.
The hit rate for materialized pages was ca. 40%.
Curiously, IOPS to the Instance Store NVMe were very low, rarely exceeding 100.
The fact that the IOPS were so low / the materialized page cache hit rate was so high suggests that **this tenant's compute's caches were thrashing**.
The compute was of type `k8s-pod`; hence, auto-scaling could/would not have helped remediate the thrashing by provisioning more RAM.
The consequence was that the **thrashing translated into excessive GetPage requests against Pageserver**.
My claim is that it was **unhealthy to serve this workload at the pace we did**:
* it is likely that other tenants were/would have experienced high latencies (again, we sadly don't have per-tenant latency data to confirm this)
* more importantly, it was **unsustainable** to serve traffic at this pace for multiple reasons:
* **predictability of performance**: when the working set grows, the pageserver materialized page cache hit rate drops.
At some point, we're bound by the EC2 Instance Store NVMe drive's IOPS limit.
The result is an **uneven** performance profile from the Compute perspective.
* **economics**: Neon currently does not charge for IOPS, only capacity.
**We cannot afford to undercut the market in IOPS/$ this drastically; it leads to adverse selection and perverse incentives.**
For example, the 155k IOPS, which we served for 10min, would cost ca. 6.5k$/month when provisioned as an io2 EBS volume.
Even the 18k IOPS, which we served for 9h, would cost ca. 1.1k$/month when provisioned as an io2 EBS volume.
We charge 0$.
It could be economically advantageous to keep using a low-DRAM compute because Pageserver IOPS are fast enough and free.
Note: It is helpful to think of Pageserver as a disk, because it's precisely where `neon_smgr` sits:
vanilla Postgres gets its pages from disk, Neon Postgres gets them from Pageserver.
So, regarding the above performance & economic arguments, it is fair to say that we currently provide an "as-fast-as-possible-IOPS" disk that we charge for only by capacity.
## Solution: Throttling GetPage Requests
**The consequence of the above analysis must be that Pageserver throttles GetPage@LSN requests**.
That is, unless we want to start charging for provisioned GetPage@LSN/second.
Throttling sets the correct incentive for a thrashing Compute to scale up its DRAM to the working set size.
Neon Autoscaling will make this easy, [eventually](https://github.com/neondatabase/neon/pull/3913).
## The Design Space
What that remains is the question about *policy* and *mechanism*:
**Policy** concerns itself with the question of what limit applies to a given connection|timeline|tenant.
Candidates are:
* hard limit, same limit value per connection|timeline|tenant
* Per-tenant will provide an upper bound for the impact of a tenant on a given Pageserver instance.
This is a major operational pain point / risk right now.
* hard limit, configurable per connection|timeline|tenant
* This outsources policy to console/control plane, with obvious advantages for flexible structuring of what service we offer to customers.
* Note that this is not a mechanism to guarantee a minium provisioned rate, i.e., this is not a mechanism to guarantee a certain QoS for a tenant.
* fair share among active connections|timelines|tenants per instance
* example: each connection|timeline|tenant gets a fair fraction of the machine's GetPage/second capacity
* NB: needs definition of "active", and knowledge of available GetPage/second capacity in advance
* ...
Regarding **mechanism**, it's clear that **backpressure** is the way to go.
However, we must choose between
* **implicit** backpressure through pq/TCP and
* **explicit** rejection of requests + retries with exponential backoff
Further, there is the question of how throttling GetPage@LSN will affect the **internal GetPage latency SLO**:
where do we measure the SLI for Pageserver's internal getpage latency SLO? Before or after the throttling?
And when we eventually move the measurement point into the Computes (to avoid coordinated omission),
how do we avoid counting throttling-induced latency toward the internal getpage latency SLI/SLO?
## Scope Of This RFC
**This RFC proposes introducing a hard GetPage@LSN/second limit per tenant, with the same value applying to each tenant on a Pageserver**.
This proposal is easy to implement and significantly de-risks operating large Pageservers,
based on the assumption that extremely-high-GetPage-rate-episodes like the one from the "Motivation" section are uncorrelated between tenants.
For example, suppose we pick a limit that allows up to 10 tenants to go at limit rate.
Suppose our Pageserver can serve 100k GetPage/second total at a 100% page cache miss rate.
If each tenant gets a hard limit of 10k GetPage/second, we can serve up to 10 tenants at limit speed without latency degradation.
The mechanism for backpressure will be TCP-based implicit backpressure.
The compute team isn't concerned about prefetch queue depth.
Pageserver will implement it by delaying the reading of requests from the libpq connection(s).
The rate limit will be implemented using a per-tenant token bucket.
The bucket will be be shared among all connections to the tenant.
The bucket implementation supports starvation-preventing `await`ing.
The current candidate for the implementation is [`leaky_bucket`](https://docs.rs/leaky-bucket/).
The getpage@lsn benchmark that's being added in https://github.com/neondatabase/neon/issues/5771
can be used to evaluate the overhead of sharing the bucket among connections of a tenant.
A possible technique to mitigate the impact of sharing the bucket would be to maintain a buffer of a few tokens per connection handler.
Regarding metrics / the internal GetPage latency SLO:
we will measure the GetPage latency SLO _after_ the throttler and introduce a new metric to measure the amount of throttling, quantified by:
- histogram that records the tenants' observations of queue depth before they start waiting (one such histogram per pageserver)
- histogram that records the tenants' observations of time spent waiting (one such histogram per pageserver)
Further observability measures:
- an INFO log message at frequency 1/min if the tenant/timeline/connection was throttled in that last minute.
The message will identify the tenant/timeline/connection to allow correlation with compute logs/stats.
Rollout will happen as follows:
- deploy 1: implementation + config: disabled by default, ability to enable it per tenant through tenant_conf
- experimentation in staging and later production to study impact & interaction with auto-scaling
- determination of a sensible global default value
- the value will be chosen as high as possible ...
- ... but low enough to work towards this RFC's goal that one tenant should not be able to dominate a pageserver instance.
- deploy 2: implementation fixes if any + config: enabled by default with the aforementioned global default
- reset of the experimental per-tenant overrides
- gain experience & lower the limit over time
- we stop lowering the limit as soon as this RFC's goal is achieved, i.e.,
once we decide that in practice the chosen value sufficiently de-risks operating large pageservers
The per-tenant override will remain for emergencies and testing.
But since Console doesn't preserve it during tenant migrations, it isn't durably configurable for the tenant.
Toward the upper layers of the Neon stack, the resulting limit will be
**"the highest GetPage/second that Pageserver can support for a single tenant"**.
### Rationale
We decided against error + retry because of worries about starvation.
## Future Work
Enable per-tenant emergency override of the limit via Console.
Should be part of a more general framework to specify tenant config overrides.
**NB:** this is **not** the right mechanism to _sell_ different max GetPage/second levels to users,
or _auto-scale_ the GetPage/second levels. Such functionality will require a separate RFC that
concerns itself with GetPage/second capacity planning.
Compute-side metrics for GetPage latency.
Back-channel to inform Compute/Autoscaling/ControlPlane that the project is being throttled.
Compute-side neon_smgr improvements to avoid sending the same GetPage request multiple times if multiple backends experience a cache miss.
Dealing with read-only endpoints: users use read-only endpoints to scale reads for a single tenant.
Possibly there are also assumptions around read-only endpoints not affecting the primary read-write endpoint's performance.
With per-tenant rate limiting, we will not meet that expectation.
However, we can currently only scale per tenant.
Soon, we will have sharding (#5505), which will apply the throttling on a per-shard basis.
But, that's orthogonal to scaling reads: if many endpoints hit one shard, they share the same throttling limit.
To solve this properly, I think we'll need replicas for tenants / shard.
To performance-isolate a tenant's endpoints from each other, we'd then route them to different replicas.

View File

@@ -2,6 +2,7 @@ pub mod partitioning;
use std::{
collections::HashMap,
io::Read,
num::{NonZeroU64, NonZeroUsize},
time::SystemTime,
};
@@ -19,7 +20,7 @@ use utils::{
use crate::{reltag::RelTag, shard::TenantShardId};
use anyhow::bail;
use bytes::{BufMut, Bytes, BytesMut};
use bytes::{Buf, BufMut, Bytes, BytesMut};
/// The state of a tenant in this pageserver.
///
@@ -369,6 +370,14 @@ pub struct TenantInfo {
pub attachment_status: TenantAttachmentStatus,
}
#[derive(Serialize, Deserialize, Clone)]
pub struct TenantDetails {
#[serde(flatten)]
pub tenant_info: TenantInfo,
pub timelines: Vec<TimelineId>,
}
/// This represents the output of the "timeline_detail" and "timeline_list" API calls.
#[derive(Debug, Serialize, Deserialize, Clone)]
pub struct TimelineInfo {
@@ -576,6 +585,7 @@ pub enum PagestreamFeMessage {
}
// Wrapped in libpq CopyData
#[derive(strum_macros::EnumProperty)]
pub enum PagestreamBeMessage {
Exists(PagestreamExistsResponse),
Nblocks(PagestreamNblocksResponse),
@@ -584,6 +594,29 @@ pub enum PagestreamBeMessage {
DbSize(PagestreamDbSizeResponse),
}
// Keep in sync with `pagestore_client.h`
#[repr(u8)]
enum PagestreamBeMessageTag {
Exists = 100,
Nblocks = 101,
GetPage = 102,
Error = 103,
DbSize = 104,
}
impl TryFrom<u8> for PagestreamBeMessageTag {
type Error = u8;
fn try_from(value: u8) -> Result<Self, u8> {
match value {
100 => Ok(PagestreamBeMessageTag::Exists),
101 => Ok(PagestreamBeMessageTag::Nblocks),
102 => Ok(PagestreamBeMessageTag::GetPage),
103 => Ok(PagestreamBeMessageTag::Error),
104 => Ok(PagestreamBeMessageTag::DbSize),
_ => Err(value),
}
}
}
#[derive(Debug, PartialEq, Eq)]
pub struct PagestreamExistsRequest {
pub latest: bool,
@@ -739,35 +772,91 @@ impl PagestreamBeMessage {
pub fn serialize(&self) -> Bytes {
let mut bytes = BytesMut::new();
use PagestreamBeMessageTag as Tag;
match self {
Self::Exists(resp) => {
bytes.put_u8(100); /* tag from pagestore_client.h */
bytes.put_u8(Tag::Exists as u8);
bytes.put_u8(resp.exists as u8);
}
Self::Nblocks(resp) => {
bytes.put_u8(101); /* tag from pagestore_client.h */
bytes.put_u8(Tag::Nblocks as u8);
bytes.put_u32(resp.n_blocks);
}
Self::GetPage(resp) => {
bytes.put_u8(102); /* tag from pagestore_client.h */
bytes.put_u8(Tag::GetPage as u8);
bytes.put(&resp.page[..]);
}
Self::Error(resp) => {
bytes.put_u8(103); /* tag from pagestore_client.h */
bytes.put_u8(Tag::Error as u8);
bytes.put(resp.message.as_bytes());
bytes.put_u8(0); // null terminator
}
Self::DbSize(resp) => {
bytes.put_u8(104); /* tag from pagestore_client.h */
bytes.put_u8(Tag::DbSize as u8);
bytes.put_i64(resp.db_size);
}
}
bytes.into()
}
pub fn deserialize(buf: Bytes) -> anyhow::Result<Self> {
let mut buf = buf.reader();
let msg_tag = buf.read_u8()?;
use PagestreamBeMessageTag as Tag;
let ok =
match Tag::try_from(msg_tag).map_err(|tag: u8| anyhow::anyhow!("invalid tag {tag}"))? {
Tag::Exists => {
let exists = buf.read_u8()?;
Self::Exists(PagestreamExistsResponse {
exists: exists != 0,
})
}
Tag::Nblocks => {
let n_blocks = buf.read_u32::<BigEndian>()?;
Self::Nblocks(PagestreamNblocksResponse { n_blocks })
}
Tag::GetPage => {
let mut page = vec![0; 8192]; // TODO: use MaybeUninit
buf.read_exact(&mut page)?;
PagestreamBeMessage::GetPage(PagestreamGetPageResponse { page: page.into() })
}
Tag::Error => {
let buf = buf.get_ref();
let cstr = std::ffi::CStr::from_bytes_until_nul(buf)?;
let rust_str = cstr.to_str()?;
PagestreamBeMessage::Error(PagestreamErrorResponse {
message: rust_str.to_owned(),
})
}
Tag::DbSize => {
let db_size = buf.read_i64::<BigEndian>()?;
Self::DbSize(PagestreamDbSizeResponse { db_size })
}
};
let remaining = buf.into_inner();
if !remaining.is_empty() {
anyhow::bail!(
"remaining bytes in msg with tag={msg_tag}: {}",
remaining.len()
);
}
Ok(ok)
}
pub fn kind(&self) -> &'static str {
match self {
Self::Exists(_) => "Exists",
Self::Nblocks(_) => "Nblocks",
Self::GetPage(_) => "GetPage",
Self::Error(_) => "Error",
Self::DbSize(_) => "DbSize",
}
}
}
#[cfg(test)]

View File

@@ -0,0 +1,200 @@
use std::collections::HashSet;
use std::ops::ControlFlow;
use std::path::PathBuf;
use std::sync::Arc;
use anyhow::Context;
use bytes::Bytes;
use camino::Utf8Path;
use futures::stream::Stream;
use once_cell::sync::OnceCell;
use remote_storage::{Download, GenericRemoteStorage, RemotePath};
use tokio::task::JoinSet;
use tracing::{debug, error, info};
static LOGGING_DONE: OnceCell<()> = OnceCell::new();
pub(crate) fn upload_stream(
content: std::borrow::Cow<'static, [u8]>,
) -> (
impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
usize,
) {
use std::borrow::Cow;
let content = match content {
Cow::Borrowed(x) => Bytes::from_static(x),
Cow::Owned(vec) => Bytes::from(vec),
};
wrap_stream(content)
}
pub(crate) fn wrap_stream(
content: bytes::Bytes,
) -> (
impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
usize,
) {
let len = content.len();
let content = futures::future::ready(Ok(content));
(futures::stream::once(content), len)
}
pub(crate) async fn download_to_vec(dl: Download) -> anyhow::Result<Vec<u8>> {
let mut buf = Vec::new();
tokio::io::copy_buf(
&mut tokio_util::io::StreamReader::new(dl.download_stream),
&mut buf,
)
.await?;
Ok(buf)
}
// Uploads files `folder{j}/blob{i}.txt`. See test description for more details.
pub(crate) async fn upload_simple_remote_data(
client: &Arc<GenericRemoteStorage>,
upload_tasks_count: usize,
) -> ControlFlow<HashSet<RemotePath>, HashSet<RemotePath>> {
info!("Creating {upload_tasks_count} remote files");
let mut upload_tasks = JoinSet::new();
for i in 1..upload_tasks_count + 1 {
let task_client = Arc::clone(client);
upload_tasks.spawn(async move {
let blob_path = PathBuf::from(format!("folder{}/blob_{}.txt", i / 7, i));
let blob_path = RemotePath::new(
Utf8Path::from_path(blob_path.as_path()).expect("must be valid blob path"),
)
.with_context(|| format!("{blob_path:?} to RemotePath conversion"))?;
debug!("Creating remote item {i} at path {blob_path:?}");
let (data, len) = upload_stream(format!("remote blob data {i}").into_bytes().into());
task_client.upload(data, len, &blob_path, None).await?;
Ok::<_, anyhow::Error>(blob_path)
});
}
let mut upload_tasks_failed = false;
let mut uploaded_blobs = HashSet::with_capacity(upload_tasks_count);
while let Some(task_run_result) = upload_tasks.join_next().await {
match task_run_result
.context("task join failed")
.and_then(|task_result| task_result.context("upload task failed"))
{
Ok(upload_path) => {
uploaded_blobs.insert(upload_path);
}
Err(e) => {
error!("Upload task failed: {e:?}");
upload_tasks_failed = true;
}
}
}
if upload_tasks_failed {
ControlFlow::Break(uploaded_blobs)
} else {
ControlFlow::Continue(uploaded_blobs)
}
}
pub(crate) async fn cleanup(
client: &Arc<GenericRemoteStorage>,
objects_to_delete: HashSet<RemotePath>,
) {
info!(
"Removing {} objects from the remote storage during cleanup",
objects_to_delete.len()
);
let mut delete_tasks = JoinSet::new();
for object_to_delete in objects_to_delete {
let task_client = Arc::clone(client);
delete_tasks.spawn(async move {
debug!("Deleting remote item at path {object_to_delete:?}");
task_client
.delete(&object_to_delete)
.await
.with_context(|| format!("{object_to_delete:?} removal"))
});
}
while let Some(task_run_result) = delete_tasks.join_next().await {
match task_run_result {
Ok(task_result) => match task_result {
Ok(()) => {}
Err(e) => error!("Delete task failed: {e:?}"),
},
Err(join_err) => error!("Delete task did not finish correctly: {join_err}"),
}
}
}
pub(crate) struct Uploads {
pub(crate) prefixes: HashSet<RemotePath>,
pub(crate) blobs: HashSet<RemotePath>,
}
pub(crate) async fn upload_remote_data(
client: &Arc<GenericRemoteStorage>,
base_prefix_str: &'static str,
upload_tasks_count: usize,
) -> ControlFlow<Uploads, Uploads> {
info!("Creating {upload_tasks_count} remote files");
let mut upload_tasks = JoinSet::new();
for i in 1..upload_tasks_count + 1 {
let task_client = Arc::clone(client);
upload_tasks.spawn(async move {
let prefix = format!("{base_prefix_str}/sub_prefix_{i}/");
let blob_prefix = RemotePath::new(Utf8Path::new(&prefix))
.with_context(|| format!("{prefix:?} to RemotePath conversion"))?;
let blob_path = blob_prefix.join(Utf8Path::new(&format!("blob_{i}")));
debug!("Creating remote item {i} at path {blob_path:?}");
let (data, data_len) =
upload_stream(format!("remote blob data {i}").into_bytes().into());
task_client.upload(data, data_len, &blob_path, None).await?;
Ok::<_, anyhow::Error>((blob_prefix, blob_path))
});
}
let mut upload_tasks_failed = false;
let mut uploaded_prefixes = HashSet::with_capacity(upload_tasks_count);
let mut uploaded_blobs = HashSet::with_capacity(upload_tasks_count);
while let Some(task_run_result) = upload_tasks.join_next().await {
match task_run_result
.context("task join failed")
.and_then(|task_result| task_result.context("upload task failed"))
{
Ok((upload_prefix, upload_path)) => {
uploaded_prefixes.insert(upload_prefix);
uploaded_blobs.insert(upload_path);
}
Err(e) => {
error!("Upload task failed: {e:?}");
upload_tasks_failed = true;
}
}
}
let uploads = Uploads {
prefixes: uploaded_prefixes,
blobs: uploaded_blobs,
};
if upload_tasks_failed {
ControlFlow::Break(uploads)
} else {
ControlFlow::Continue(uploads)
}
}
pub(crate) fn ensure_logging_ready() {
LOGGING_DONE.get_or_init(|| {
utils::logging::init(
utils::logging::LogFormat::Test,
utils::logging::TracingErrorLayerEnablement::Disabled,
utils::logging::Output::Stdout,
)
.expect("logging init failed");
});
}

View File

@@ -2,23 +2,23 @@ use std::collections::HashSet;
use std::env;
use std::num::NonZeroUsize;
use std::ops::ControlFlow;
use std::path::PathBuf;
use std::sync::Arc;
use std::time::UNIX_EPOCH;
use anyhow::Context;
use bytes::Bytes;
use camino::Utf8Path;
use futures::stream::Stream;
use once_cell::sync::OnceCell;
use remote_storage::{
AzureConfig, Download, GenericRemoteStorage, RemotePath, RemoteStorageConfig, RemoteStorageKind,
AzureConfig, GenericRemoteStorage, RemotePath, RemoteStorageConfig, RemoteStorageKind,
};
use test_context::{test_context, AsyncTestContext};
use tokio::task::JoinSet;
use tracing::{debug, error, info};
use tracing::{debug, info};
static LOGGING_DONE: OnceCell<()> = OnceCell::new();
mod common;
use common::{
cleanup, download_to_vec, ensure_logging_ready, upload_remote_data, upload_simple_remote_data,
upload_stream, wrap_stream,
};
const ENABLE_REAL_AZURE_REMOTE_STORAGE_ENV_VAR_NAME: &str = "ENABLE_REAL_AZURE_REMOTE_STORAGE";
@@ -30,7 +30,7 @@ const BASE_PREFIX: &str = "test";
/// If real Azure tests are disabled, the test passes, skipping any real test run: currently, there's no way to mark the test ignored in runtime with the
/// deafult test framework, see https://github.com/rust-lang/rust/issues/68007 for details.
///
/// First, the test creates a set of Azure blobs with keys `/${random_prefix_part}/${base_prefix_str}/sub_prefix_${i}/blob_${i}` in [`upload_azure_data`]
/// First, the test creates a set of Azure blobs with keys `/${random_prefix_part}/${base_prefix_str}/sub_prefix_${i}/blob_${i}` in [`upload_remote_data`]
/// where
/// * `random_prefix_part` is set for the entire Azure client during the Azure client creation in [`create_azure_client`], to avoid multiple test runs interference
/// * `base_prefix_str` is a common prefix to use in the client requests: we would want to ensure that the client is able to list nested prefixes inside the bucket
@@ -97,7 +97,7 @@ async fn azure_pagination_should_work(
/// Uses real Azure and requires [`ENABLE_REAL_AZURE_REMOTE_STORAGE_ENV_VAR_NAME`] and related Azure cred env vars specified. Test will skip real code and pass if env vars not set.
/// See `Azure_pagination_should_work` for more information.
///
/// First, create a set of Azure objects with keys `random_prefix/folder{j}/blob_{i}.txt` in [`upload_azure_data`]
/// First, create a set of Azure objects with keys `random_prefix/folder{j}/blob_{i}.txt` in [`upload_remote_data`]
/// Then performs the following queries:
/// 1. `list_files(None)`. This should return all files `random_prefix/folder{j}/blob_{i}.txt`
/// 2. `list_files("folder1")`. This should return all files `random_prefix/folder1/blob_{i}.txt`
@@ -218,18 +218,9 @@ async fn azure_upload_download_works(ctx: &mut MaybeEnabledAzure) -> anyhow::Res
ctx.client.upload(data, len, &path, None).await?;
async fn download_and_compare(dl: Download) -> anyhow::Result<Vec<u8>> {
let mut buf = Vec::new();
tokio::io::copy_buf(
&mut tokio_util::io::StreamReader::new(dl.download_stream),
&mut buf,
)
.await?;
Ok(buf)
}
// Normal download request
let dl = ctx.client.download(&path).await?;
let buf = download_and_compare(dl).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig);
// Full range (end specified)
@@ -237,12 +228,12 @@ async fn azure_upload_download_works(ctx: &mut MaybeEnabledAzure) -> anyhow::Res
.client
.download_byte_range(&path, 0, Some(len as u64))
.await?;
let buf = download_and_compare(dl).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig);
// partial range (end specified)
let dl = ctx.client.download_byte_range(&path, 4, Some(10)).await?;
let buf = download_and_compare(dl).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig[4..10]);
// partial range (end beyond real end)
@@ -250,17 +241,17 @@ async fn azure_upload_download_works(ctx: &mut MaybeEnabledAzure) -> anyhow::Res
.client
.download_byte_range(&path, 8, Some(len as u64 * 100))
.await?;
let buf = download_and_compare(dl).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig[8..]);
// Partial range (end unspecified)
let dl = ctx.client.download_byte_range(&path, 4, None).await?;
let buf = download_and_compare(dl).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig[4..]);
// Full range (end unspecified)
let dl = ctx.client.download_byte_range(&path, 0, None).await?;
let buf = download_and_compare(dl).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig);
debug!("Cleanup: deleting file at path {path:?}");
@@ -272,17 +263,6 @@ async fn azure_upload_download_works(ctx: &mut MaybeEnabledAzure) -> anyhow::Res
Ok(())
}
fn ensure_logging_ready() {
LOGGING_DONE.get_or_init(|| {
utils::logging::init(
utils::logging::LogFormat::Test,
utils::logging::TracingErrorLayerEnablement::Disabled,
utils::logging::Output::Stdout,
)
.expect("logging init failed");
});
}
struct EnabledAzure {
client: Arc<GenericRemoteStorage>,
base_prefix: &'static str,
@@ -352,7 +332,7 @@ impl AsyncTestContext for MaybeEnabledAzureWithTestBlobs {
let enabled = EnabledAzure::setup(Some(max_keys_in_list_response)).await;
match upload_azure_data(&enabled.client, enabled.base_prefix, upload_tasks_count).await {
match upload_remote_data(&enabled.client, enabled.base_prefix, upload_tasks_count).await {
ControlFlow::Continue(uploads) => {
info!("Remote objects created successfully");
@@ -414,7 +394,7 @@ impl AsyncTestContext for MaybeEnabledAzureWithSimpleTestBlobs {
let enabled = EnabledAzure::setup(Some(max_keys_in_list_response)).await;
match upload_simple_azure_data(&enabled.client, upload_tasks_count).await {
match upload_simple_remote_data(&enabled.client, upload_tasks_count).await {
ControlFlow::Continue(uploads) => {
info!("Remote objects created successfully");
@@ -478,166 +458,3 @@ fn create_azure_client(
GenericRemoteStorage::from_config(&remote_storage_config).context("remote storage init")?,
))
}
struct Uploads {
prefixes: HashSet<RemotePath>,
blobs: HashSet<RemotePath>,
}
async fn upload_azure_data(
client: &Arc<GenericRemoteStorage>,
base_prefix_str: &'static str,
upload_tasks_count: usize,
) -> ControlFlow<Uploads, Uploads> {
info!("Creating {upload_tasks_count} Azure files");
let mut upload_tasks = JoinSet::new();
for i in 1..upload_tasks_count + 1 {
let task_client = Arc::clone(client);
upload_tasks.spawn(async move {
let prefix = format!("{base_prefix_str}/sub_prefix_{i}/");
let blob_prefix = RemotePath::new(Utf8Path::new(&prefix))
.with_context(|| format!("{prefix:?} to RemotePath conversion"))?;
let blob_path = blob_prefix.join(Utf8Path::new(&format!("blob_{i}")));
debug!("Creating remote item {i} at path {blob_path:?}");
let (data, len) = upload_stream(format!("remote blob data {i}").into_bytes().into());
task_client.upload(data, len, &blob_path, None).await?;
Ok::<_, anyhow::Error>((blob_prefix, blob_path))
});
}
let mut upload_tasks_failed = false;
let mut uploaded_prefixes = HashSet::with_capacity(upload_tasks_count);
let mut uploaded_blobs = HashSet::with_capacity(upload_tasks_count);
while let Some(task_run_result) = upload_tasks.join_next().await {
match task_run_result
.context("task join failed")
.and_then(|task_result| task_result.context("upload task failed"))
{
Ok((upload_prefix, upload_path)) => {
uploaded_prefixes.insert(upload_prefix);
uploaded_blobs.insert(upload_path);
}
Err(e) => {
error!("Upload task failed: {e:?}");
upload_tasks_failed = true;
}
}
}
let uploads = Uploads {
prefixes: uploaded_prefixes,
blobs: uploaded_blobs,
};
if upload_tasks_failed {
ControlFlow::Break(uploads)
} else {
ControlFlow::Continue(uploads)
}
}
async fn cleanup(client: &Arc<GenericRemoteStorage>, objects_to_delete: HashSet<RemotePath>) {
info!(
"Removing {} objects from the remote storage during cleanup",
objects_to_delete.len()
);
let mut delete_tasks = JoinSet::new();
for object_to_delete in objects_to_delete {
let task_client = Arc::clone(client);
delete_tasks.spawn(async move {
debug!("Deleting remote item at path {object_to_delete:?}");
task_client
.delete(&object_to_delete)
.await
.with_context(|| format!("{object_to_delete:?} removal"))
});
}
while let Some(task_run_result) = delete_tasks.join_next().await {
match task_run_result {
Ok(task_result) => match task_result {
Ok(()) => {}
Err(e) => error!("Delete task failed: {e:?}"),
},
Err(join_err) => error!("Delete task did not finish correctly: {join_err}"),
}
}
}
// Uploads files `folder{j}/blob{i}.txt`. See test description for more details.
async fn upload_simple_azure_data(
client: &Arc<GenericRemoteStorage>,
upload_tasks_count: usize,
) -> ControlFlow<HashSet<RemotePath>, HashSet<RemotePath>> {
info!("Creating {upload_tasks_count} Azure files");
let mut upload_tasks = JoinSet::new();
for i in 1..upload_tasks_count + 1 {
let task_client = Arc::clone(client);
upload_tasks.spawn(async move {
let blob_path = PathBuf::from(format!("folder{}/blob_{}.txt", i / 7, i));
let blob_path = RemotePath::new(
Utf8Path::from_path(blob_path.as_path()).expect("must be valid blob path"),
)
.with_context(|| format!("{blob_path:?} to RemotePath conversion"))?;
debug!("Creating remote item {i} at path {blob_path:?}");
let (data, len) = upload_stream(format!("remote blob data {i}").into_bytes().into());
task_client.upload(data, len, &blob_path, None).await?;
Ok::<_, anyhow::Error>(blob_path)
});
}
let mut upload_tasks_failed = false;
let mut uploaded_blobs = HashSet::with_capacity(upload_tasks_count);
while let Some(task_run_result) = upload_tasks.join_next().await {
match task_run_result
.context("task join failed")
.and_then(|task_result| task_result.context("upload task failed"))
{
Ok(upload_path) => {
uploaded_blobs.insert(upload_path);
}
Err(e) => {
error!("Upload task failed: {e:?}");
upload_tasks_failed = true;
}
}
}
if upload_tasks_failed {
ControlFlow::Break(uploaded_blobs)
} else {
ControlFlow::Continue(uploaded_blobs)
}
}
// FIXME: copypasted from test_real_s3, can't remember how to share a module which is not compiled
// to binary
fn upload_stream(
content: std::borrow::Cow<'static, [u8]>,
) -> (
impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
usize,
) {
use std::borrow::Cow;
let content = match content {
Cow::Borrowed(x) => Bytes::from_static(x),
Cow::Owned(vec) => Bytes::from(vec),
};
wrap_stream(content)
}
fn wrap_stream(
content: bytes::Bytes,
) -> (
impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
usize,
) {
let len = content.len();
let content = futures::future::ready(Ok(content));
(futures::stream::once(content), len)
}

View File

@@ -2,23 +2,23 @@ use std::collections::HashSet;
use std::env;
use std::num::NonZeroUsize;
use std::ops::ControlFlow;
use std::path::PathBuf;
use std::sync::Arc;
use std::time::UNIX_EPOCH;
use anyhow::Context;
use bytes::Bytes;
use camino::Utf8Path;
use futures::stream::Stream;
use once_cell::sync::OnceCell;
use remote_storage::{
GenericRemoteStorage, RemotePath, RemoteStorageConfig, RemoteStorageKind, S3Config,
};
use test_context::{test_context, AsyncTestContext};
use tokio::task::JoinSet;
use tracing::{debug, error, info};
use tracing::{debug, info};
static LOGGING_DONE: OnceCell<()> = OnceCell::new();
mod common;
use common::{
cleanup, download_to_vec, ensure_logging_ready, upload_remote_data, upload_simple_remote_data,
upload_stream, wrap_stream,
};
const ENABLE_REAL_S3_REMOTE_STORAGE_ENV_VAR_NAME: &str = "ENABLE_REAL_S3_REMOTE_STORAGE";
@@ -30,7 +30,7 @@ const BASE_PREFIX: &str = "test";
/// If real S3 tests are disabled, the test passes, skipping any real test run: currently, there's no way to mark the test ignored in runtime with the
/// deafult test framework, see https://github.com/rust-lang/rust/issues/68007 for details.
///
/// First, the test creates a set of S3 objects with keys `/${random_prefix_part}/${base_prefix_str}/sub_prefix_${i}/blob_${i}` in [`upload_s3_data`]
/// First, the test creates a set of S3 objects with keys `/${random_prefix_part}/${base_prefix_str}/sub_prefix_${i}/blob_${i}` in [`upload_remote_data`]
/// where
/// * `random_prefix_part` is set for the entire S3 client during the S3 client creation in [`create_s3_client`], to avoid multiple test runs interference
/// * `base_prefix_str` is a common prefix to use in the client requests: we would want to ensure that the client is able to list nested prefixes inside the bucket
@@ -95,7 +95,7 @@ async fn s3_pagination_should_work(ctx: &mut MaybeEnabledS3WithTestBlobs) -> any
/// Uses real S3 and requires [`ENABLE_REAL_S3_REMOTE_STORAGE_ENV_VAR_NAME`] and related S3 cred env vars specified. Test will skip real code and pass if env vars not set.
/// See `s3_pagination_should_work` for more information.
///
/// First, create a set of S3 objects with keys `random_prefix/folder{j}/blob_{i}.txt` in [`upload_s3_data`]
/// First, create a set of S3 objects with keys `random_prefix/folder{j}/blob_{i}.txt` in [`upload_remote_data`]
/// Then performs the following queries:
/// 1. `list_files(None)`. This should return all files `random_prefix/folder{j}/blob_{i}.txt`
/// 2. `list_files("folder1")`. This should return all files `random_prefix/folder1/blob_{i}.txt`
@@ -198,15 +198,65 @@ async fn s3_delete_objects_works(ctx: &mut MaybeEnabledS3) -> anyhow::Result<()>
Ok(())
}
fn ensure_logging_ready() {
LOGGING_DONE.get_or_init(|| {
utils::logging::init(
utils::logging::LogFormat::Test,
utils::logging::TracingErrorLayerEnablement::Disabled,
utils::logging::Output::Stdout,
)
.expect("logging init failed");
});
#[test_context(MaybeEnabledS3)]
#[tokio::test]
async fn s3_upload_download_works(ctx: &mut MaybeEnabledS3) -> anyhow::Result<()> {
let MaybeEnabledS3::Enabled(ctx) = ctx else {
return Ok(());
};
let path = RemotePath::new(Utf8Path::new(format!("{}/file", ctx.base_prefix).as_str()))
.with_context(|| "RemotePath conversion")?;
let orig = bytes::Bytes::from_static("remote blob data here".as_bytes());
let (data, len) = wrap_stream(orig.clone());
ctx.client.upload(data, len, &path, None).await?;
// Normal download request
let dl = ctx.client.download(&path).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig);
// Full range (end specified)
let dl = ctx
.client
.download_byte_range(&path, 0, Some(len as u64))
.await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig);
// partial range (end specified)
let dl = ctx.client.download_byte_range(&path, 4, Some(10)).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig[4..10]);
// partial range (end beyond real end)
let dl = ctx
.client
.download_byte_range(&path, 8, Some(len as u64 * 100))
.await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig[8..]);
// Partial range (end unspecified)
let dl = ctx.client.download_byte_range(&path, 4, None).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig[4..]);
// Full range (end unspecified)
let dl = ctx.client.download_byte_range(&path, 0, None).await?;
let buf = download_to_vec(dl).await?;
assert_eq!(&buf, &orig);
debug!("Cleanup: deleting file at path {path:?}");
ctx.client
.delete(&path)
.await
.with_context(|| format!("{path:?} removal"))?;
Ok(())
}
struct EnabledS3 {
@@ -278,7 +328,7 @@ impl AsyncTestContext for MaybeEnabledS3WithTestBlobs {
let enabled = EnabledS3::setup(Some(max_keys_in_list_response)).await;
match upload_s3_data(&enabled.client, enabled.base_prefix, upload_tasks_count).await {
match upload_remote_data(&enabled.client, enabled.base_prefix, upload_tasks_count).await {
ControlFlow::Continue(uploads) => {
info!("Remote objects created successfully");
@@ -340,7 +390,7 @@ impl AsyncTestContext for MaybeEnabledS3WithSimpleTestBlobs {
let enabled = EnabledS3::setup(Some(max_keys_in_list_response)).await;
match upload_simple_s3_data(&enabled.client, upload_tasks_count).await {
match upload_simple_remote_data(&enabled.client, upload_tasks_count).await {
ControlFlow::Continue(uploads) => {
info!("Remote objects created successfully");
@@ -403,166 +453,3 @@ fn create_s3_client(
GenericRemoteStorage::from_config(&remote_storage_config).context("remote storage init")?,
))
}
struct Uploads {
prefixes: HashSet<RemotePath>,
blobs: HashSet<RemotePath>,
}
async fn upload_s3_data(
client: &Arc<GenericRemoteStorage>,
base_prefix_str: &'static str,
upload_tasks_count: usize,
) -> ControlFlow<Uploads, Uploads> {
info!("Creating {upload_tasks_count} S3 files");
let mut upload_tasks = JoinSet::new();
for i in 1..upload_tasks_count + 1 {
let task_client = Arc::clone(client);
upload_tasks.spawn(async move {
let prefix = format!("{base_prefix_str}/sub_prefix_{i}/");
let blob_prefix = RemotePath::new(Utf8Path::new(&prefix))
.with_context(|| format!("{prefix:?} to RemotePath conversion"))?;
let blob_path = blob_prefix.join(Utf8Path::new(&format!("blob_{i}")));
debug!("Creating remote item {i} at path {blob_path:?}");
let (data, data_len) =
upload_stream(format!("remote blob data {i}").into_bytes().into());
task_client.upload(data, data_len, &blob_path, None).await?;
Ok::<_, anyhow::Error>((blob_prefix, blob_path))
});
}
let mut upload_tasks_failed = false;
let mut uploaded_prefixes = HashSet::with_capacity(upload_tasks_count);
let mut uploaded_blobs = HashSet::with_capacity(upload_tasks_count);
while let Some(task_run_result) = upload_tasks.join_next().await {
match task_run_result
.context("task join failed")
.and_then(|task_result| task_result.context("upload task failed"))
{
Ok((upload_prefix, upload_path)) => {
uploaded_prefixes.insert(upload_prefix);
uploaded_blobs.insert(upload_path);
}
Err(e) => {
error!("Upload task failed: {e:?}");
upload_tasks_failed = true;
}
}
}
let uploads = Uploads {
prefixes: uploaded_prefixes,
blobs: uploaded_blobs,
};
if upload_tasks_failed {
ControlFlow::Break(uploads)
} else {
ControlFlow::Continue(uploads)
}
}
async fn cleanup(client: &Arc<GenericRemoteStorage>, objects_to_delete: HashSet<RemotePath>) {
info!(
"Removing {} objects from the remote storage during cleanup",
objects_to_delete.len()
);
let mut delete_tasks = JoinSet::new();
for object_to_delete in objects_to_delete {
let task_client = Arc::clone(client);
delete_tasks.spawn(async move {
debug!("Deleting remote item at path {object_to_delete:?}");
task_client
.delete(&object_to_delete)
.await
.with_context(|| format!("{object_to_delete:?} removal"))
});
}
while let Some(task_run_result) = delete_tasks.join_next().await {
match task_run_result {
Ok(task_result) => match task_result {
Ok(()) => {}
Err(e) => error!("Delete task failed: {e:?}"),
},
Err(join_err) => error!("Delete task did not finish correctly: {join_err}"),
}
}
}
// Uploads files `folder{j}/blob{i}.txt`. See test description for more details.
async fn upload_simple_s3_data(
client: &Arc<GenericRemoteStorage>,
upload_tasks_count: usize,
) -> ControlFlow<HashSet<RemotePath>, HashSet<RemotePath>> {
info!("Creating {upload_tasks_count} S3 files");
let mut upload_tasks = JoinSet::new();
for i in 1..upload_tasks_count + 1 {
let task_client = Arc::clone(client);
upload_tasks.spawn(async move {
let blob_path = PathBuf::from(format!("folder{}/blob_{}.txt", i / 7, i));
let blob_path = RemotePath::new(
Utf8Path::from_path(blob_path.as_path()).expect("must be valid blob path"),
)
.with_context(|| format!("{blob_path:?} to RemotePath conversion"))?;
debug!("Creating remote item {i} at path {blob_path:?}");
let (data, data_len) =
upload_stream(format!("remote blob data {i}").into_bytes().into());
task_client.upload(data, data_len, &blob_path, None).await?;
Ok::<_, anyhow::Error>(blob_path)
});
}
let mut upload_tasks_failed = false;
let mut uploaded_blobs = HashSet::with_capacity(upload_tasks_count);
while let Some(task_run_result) = upload_tasks.join_next().await {
match task_run_result
.context("task join failed")
.and_then(|task_result| task_result.context("upload task failed"))
{
Ok(upload_path) => {
uploaded_blobs.insert(upload_path);
}
Err(e) => {
error!("Upload task failed: {e:?}");
upload_tasks_failed = true;
}
}
}
if upload_tasks_failed {
ControlFlow::Break(uploaded_blobs)
} else {
ControlFlow::Continue(uploaded_blobs)
}
}
fn upload_stream(
content: std::borrow::Cow<'static, [u8]>,
) -> (
impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
usize,
) {
use std::borrow::Cow;
let content = match content {
Cow::Borrowed(x) => Bytes::from_static(x),
Cow::Owned(vec) => Bytes::from(vec),
};
wrap_stream(content)
}
fn wrap_stream(
content: bytes::Bytes,
) -> (
impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
usize,
) {
let len = content.len();
let content = futures::future::ready(Ok(content));
(futures::stream::once(content), len)
}

View File

@@ -1 +1,2 @@
#include "postgres.h"
#include "walproposer.h"

View File

@@ -63,6 +63,7 @@ thiserror.workspace = true
tokio = { workspace = true, features = ["process", "sync", "fs", "rt", "io-util", "time"] }
tokio-io-timeout.workspace = true
tokio-postgres.workspace = true
tokio-stream.workspace = true
tokio-util.workspace = true
toml_edit = { workspace = true, features = [ "serde" ] }
tracing.workspace = true

View File

@@ -12,3 +12,11 @@ reqwest.workspace = true
utils.workspace = true
serde.workspace = true
workspace_hack = { version = "0.1", path = "../../workspace_hack" }
tokio-postgres.workspace = true
tokio-stream.workspace = true
tokio.workspace = true
futures.workspace = true
tokio-util.workspace = true
anyhow.workspace = true
postgres.workspace = true
bytes.workspace = true

View File

@@ -1 +1,2 @@
pub mod mgmt_api;
pub mod page_service;

View File

@@ -64,6 +64,18 @@ impl Client {
resp.json().await.map_err(Error::ReceiveBody)
}
pub async fn tenant_details(
&self,
tenant_id: TenantId,
) -> Result<pageserver_api::models::TenantDetails> {
let uri = format!("{}/v1/tenant/{tenant_id}", self.mgmt_api_endpoint);
self.get(uri)
.await?
.json()
.await
.map_err(Error::ReceiveBody)
}
pub async fn list_timelines(
&self,
tenant_id: TenantId,

View File

@@ -0,0 +1,151 @@
use std::pin::Pin;
use futures::SinkExt;
use pageserver_api::{
models::{
PagestreamBeMessage, PagestreamFeMessage, PagestreamGetPageRequest,
PagestreamGetPageResponse,
},
reltag::RelTag,
};
use tokio::task::JoinHandle;
use tokio_postgres::CopyOutStream;
use tokio_stream::StreamExt;
use tokio_util::sync::CancellationToken;
use utils::{
id::{TenantId, TimelineId},
lsn::Lsn,
};
pub struct Client {
client: tokio_postgres::Client,
cancel_on_client_drop: Option<tokio_util::sync::DropGuard>,
conn_task: JoinHandle<()>,
}
pub struct BasebackupRequest {
pub tenant_id: TenantId,
pub timeline_id: TimelineId,
pub lsn: Option<Lsn>,
pub gzip: bool,
}
impl Client {
pub async fn new(connstring: String) -> anyhow::Result<Self> {
let (client, connection) = tokio_postgres::connect(&connstring, postgres::NoTls).await?;
let conn_task_cancel = CancellationToken::new();
let conn_task = tokio::spawn({
let conn_task_cancel = conn_task_cancel.clone();
async move {
tokio::select! {
_ = conn_task_cancel.cancelled() => { }
res = connection => {
res.unwrap();
}
}
}
});
Ok(Self {
cancel_on_client_drop: Some(conn_task_cancel.drop_guard()),
conn_task,
client,
})
}
pub async fn pagestream(
self,
tenant_id: TenantId,
timeline_id: TimelineId,
) -> anyhow::Result<PagestreamClient> {
let copy_both: tokio_postgres::CopyBothDuplex<bytes::Bytes> = self
.client
.copy_both_simple(&format!("pagestream {tenant_id} {timeline_id}"))
.await?;
let Client {
cancel_on_client_drop,
conn_task,
client: _,
} = self;
Ok(PagestreamClient {
copy_both: Box::pin(copy_both),
conn_task,
cancel_on_client_drop,
})
}
pub async fn basebackup(&self, req: &BasebackupRequest) -> anyhow::Result<CopyOutStream> {
let BasebackupRequest {
tenant_id,
timeline_id,
lsn,
gzip,
} = req;
let mut args = Vec::with_capacity(5);
args.push("basebackup".to_string());
args.push(format!("{tenant_id}"));
args.push(format!("{timeline_id}"));
if let Some(lsn) = lsn {
args.push(format!("{lsn}"));
}
if *gzip {
args.push("--gzip".to_string())
}
Ok(self.client.copy_out(&args.join(" ")).await?)
}
}
/// Create using [`Client::pagestream`].
pub struct PagestreamClient {
copy_both: Pin<Box<tokio_postgres::CopyBothDuplex<bytes::Bytes>>>,
cancel_on_client_drop: Option<tokio_util::sync::DropGuard>,
conn_task: JoinHandle<()>,
}
pub struct RelTagBlockNo {
pub rel_tag: RelTag,
pub block_no: u32,
}
impl PagestreamClient {
pub async fn shutdown(mut self) {
let _ = self.cancel_on_client_drop.take();
self.conn_task.await.unwrap();
}
pub async fn getpage(
&mut self,
key: RelTagBlockNo,
lsn: Lsn,
) -> anyhow::Result<PagestreamGetPageResponse> {
let req = PagestreamGetPageRequest {
latest: false,
rel: key.rel_tag,
blkno: key.block_no,
lsn,
};
let req = PagestreamFeMessage::GetPage(req);
let req: bytes::Bytes = req.serialize();
// let mut req = tokio_util::io::ReaderStream::new(&req);
let mut req = tokio_stream::once(Ok(req));
self.copy_both.send_all(&mut req).await?;
let next: Option<Result<bytes::Bytes, _>> = self.copy_both.next().await;
let next: bytes::Bytes = next.unwrap()?;
let msg = PagestreamBeMessage::deserialize(next)?;
match msg {
PagestreamBeMessage::GetPage(p) => Ok(p),
PagestreamBeMessage::Error(e) => anyhow::bail!("Error: {:?}", e),
PagestreamBeMessage::Exists(_)
| PagestreamBeMessage::Nblocks(_)
| PagestreamBeMessage::DbSize(_) => {
anyhow::bail!(
"unexpected be message kind in response to getpage request: {}",
msg.kind()
)
}
}
}
}

View File

@@ -14,6 +14,7 @@ use hyper::header;
use hyper::StatusCode;
use hyper::{Body, Request, Response, Uri};
use metrics::launch_timestamp::LaunchTimestamp;
use pageserver_api::models::TenantDetails;
use pageserver_api::models::{
DownloadRemoteLayersTaskSpawnRequest, LocationConfigMode, TenantAttachRequest,
TenantLoadRequest, TenantLocationConfigRequest,
@@ -592,8 +593,6 @@ async fn get_lsn_by_timestamp_handler(
)));
}
let version: Option<u8> = parse_query_param(&request, "version")?;
let timeline_id: TimelineId = parse_request_param(&request, "timeline_id")?;
let timestamp_raw = must_get_query_param(&request, "timestamp")?;
let timestamp = humantime::parse_rfc3339(&timestamp_raw)
@@ -606,31 +605,18 @@ async fn get_lsn_by_timestamp_handler(
let result = timeline
.find_lsn_for_timestamp(timestamp_pg, &cancel, &ctx)
.await?;
if version.unwrap_or(0) > 1 {
#[derive(serde::Serialize)]
struct Result {
lsn: Lsn,
kind: &'static str,
}
let (lsn, kind) = match result {
LsnForTimestamp::Present(lsn) => (lsn, "present"),
LsnForTimestamp::Future(lsn) => (lsn, "future"),
LsnForTimestamp::Past(lsn) => (lsn, "past"),
LsnForTimestamp::NoData(lsn) => (lsn, "nodata"),
};
json_response(StatusCode::OK, Result { lsn, kind })
} else {
// FIXME: this is a temporary crutch not to break backwards compatibility
// See https://github.com/neondatabase/neon/pull/5608
let result = match result {
LsnForTimestamp::Present(lsn) => format!("{lsn}"),
LsnForTimestamp::Future(_lsn) => "future".into(),
LsnForTimestamp::Past(_lsn) => "past".into(),
LsnForTimestamp::NoData(_lsn) => "nodata".into(),
};
json_response(StatusCode::OK, result)
#[derive(serde::Serialize)]
struct Result {
lsn: Lsn,
kind: &'static str,
}
let (lsn, kind) = match result {
LsnForTimestamp::Present(lsn) => (lsn, "present"),
LsnForTimestamp::Future(lsn) => (lsn, "future"),
LsnForTimestamp::Past(lsn) => (lsn, "past"),
LsnForTimestamp::NoData(lsn) => (lsn, "nodata"),
};
json_response(StatusCode::OK, Result { lsn, kind })
}
async fn get_timestamp_of_lsn_handler(
@@ -872,11 +858,14 @@ async fn tenant_status(
}
let state = tenant.current_state();
Result::<_, ApiError>::Ok(TenantInfo {
id: tenant_shard_id,
state: state.clone(),
current_physical_size: Some(current_physical_size),
attachment_status: state.attachment_status(),
Result::<_, ApiError>::Ok(TenantDetails {
tenant_info: TenantInfo {
id: tenant_shard_id,
state: state.clone(),
current_physical_size: Some(current_physical_size),
attachment_status: state.attachment_status(),
},
timelines: tenant.list_timeline_ids(),
})
}
.instrument(info_span!("tenant_status_handler",

View File

@@ -1552,6 +1552,10 @@ impl Tenant {
.collect()
}
pub fn list_timeline_ids(&self) -> Vec<TimelineId> {
self.timelines.lock().unwrap().keys().cloned().collect()
}
/// This is used to create the initial 'main' timeline during bootstrapping,
/// or when importing a new base backup. The caller is expected to load an
/// initial image of the datadir to the new timeline after this.

View File

@@ -259,8 +259,9 @@ impl Layer {
layer
.get_value_reconstruct_data(key, lsn_range, reconstruct_data, &self.0, ctx)
.instrument(tracing::info_span!("get_value_reconstruct_data", layer=%self))
.instrument(tracing::debug_span!("get_value_reconstruct_data", layer=%self))
.await
.with_context(|| format!("get_value_reconstruct_data for layer {self}"))
}
/// Download the layer if evicted.
@@ -654,7 +655,6 @@ impl LayerInner {
}
/// Cancellation safe.
#[tracing::instrument(skip_all, fields(layer=%self))]
async fn get_or_maybe_download(
self: &Arc<Self>,
allow_download: bool,
@@ -663,95 +663,101 @@ impl LayerInner {
let mut init_permit = None;
loop {
let download = move |permit| async move {
// disable any scheduled but not yet running eviction deletions for this
let next_version = 1 + self.version.fetch_add(1, Ordering::Relaxed);
let download = move |permit| {
async move {
// disable any scheduled but not yet running eviction deletions for this
let next_version = 1 + self.version.fetch_add(1, Ordering::Relaxed);
// count cancellations, which currently remain largely unexpected
let init_cancelled =
scopeguard::guard((), |_| LAYER_IMPL_METRICS.inc_init_cancelled());
// count cancellations, which currently remain largely unexpected
let init_cancelled =
scopeguard::guard((), |_| LAYER_IMPL_METRICS.inc_init_cancelled());
// no need to make the evict_and_wait wait for the actual download to complete
drop(self.status.send(Status::Downloaded));
// no need to make the evict_and_wait wait for the actual download to complete
drop(self.status.send(Status::Downloaded));
let timeline = self
.timeline
.upgrade()
.ok_or_else(|| DownloadError::TimelineShutdown)?;
let timeline = self
.timeline
.upgrade()
.ok_or_else(|| DownloadError::TimelineShutdown)?;
// FIXME: grab a gate
// FIXME: grab a gate
let can_ever_evict = timeline.remote_client.as_ref().is_some();
let can_ever_evict = timeline.remote_client.as_ref().is_some();
// check if we really need to be downloaded; could have been already downloaded by a
// cancelled previous attempt.
let needs_download = self
.needs_download()
.await
.map_err(DownloadError::PreStatFailed)?;
// check if we really need to be downloaded; could have been already downloaded by a
// cancelled previous attempt.
let needs_download = self
.needs_download()
.await
.map_err(DownloadError::PreStatFailed)?;
let permit = if let Some(reason) = needs_download {
if let NeedsDownload::NotFile(ft) = reason {
return Err(DownloadError::NotFile(ft));
let permit = if let Some(reason) = needs_download {
if let NeedsDownload::NotFile(ft) = reason {
return Err(DownloadError::NotFile(ft));
}
// only reset this after we've decided we really need to download. otherwise it'd
// be impossible to mark cancelled downloads for eviction, like one could imagine
// we would like to do for prefetching which was not needed.
self.wanted_evicted.store(false, Ordering::Release);
if !can_ever_evict {
return Err(DownloadError::NoRemoteStorage);
}
if let Some(ctx) = ctx {
self.check_expected_download(ctx)?;
}
if !allow_download {
// this does look weird, but for LayerInner the "downloading" means also changing
// internal once related state ...
return Err(DownloadError::DownloadRequired);
}
tracing::info!(%reason, "downloading on-demand");
self.spawn_download_and_wait(timeline, permit).await?
} else {
// the file is present locally, probably by a previous but cancelled call to
// get_or_maybe_download. alternatively we might be running without remote storage.
LAYER_IMPL_METRICS.inc_init_needed_no_download();
permit
};
let since_last_eviction =
self.last_evicted_at.lock().unwrap().map(|ts| ts.elapsed());
if let Some(since_last_eviction) = since_last_eviction {
// FIXME: this will not always be recorded correctly until #6028 (the no
// download needed branch above)
LAYER_IMPL_METRICS.record_redownloaded_after(since_last_eviction);
}
// only reset this after we've decided we really need to download. otherwise it'd
// be impossible to mark cancelled downloads for eviction, like one could imagine
// we would like to do for prefetching which was not needed.
self.wanted_evicted.store(false, Ordering::Release);
let res = Arc::new(DownloadedLayer {
owner: Arc::downgrade(self),
kind: tokio::sync::OnceCell::default(),
version: next_version,
});
if !can_ever_evict {
return Err(DownloadError::NoRemoteStorage);
self.access_stats.record_residence_event(
LayerResidenceStatus::Resident,
LayerResidenceEventReason::ResidenceChange,
);
let waiters = self.inner.initializer_count();
if waiters > 0 {
tracing::info!(
waiters,
"completing the on-demand download for other tasks"
);
}
if let Some(ctx) = ctx {
self.check_expected_download(ctx)?;
}
scopeguard::ScopeGuard::into_inner(init_cancelled);
if !allow_download {
// this does look weird, but for LayerInner the "downloading" means also changing
// internal once related state ...
return Err(DownloadError::DownloadRequired);
}
tracing::info!(%reason, "downloading on-demand");
self.spawn_download_and_wait(timeline, permit).await?
} else {
// the file is present locally, probably by a previous but cancelled call to
// get_or_maybe_download. alternatively we might be running without remote storage.
LAYER_IMPL_METRICS.inc_init_needed_no_download();
permit
};
let since_last_eviction =
self.last_evicted_at.lock().unwrap().map(|ts| ts.elapsed());
if let Some(since_last_eviction) = since_last_eviction {
// FIXME: this will not always be recorded correctly until #6028 (the no
// download needed branch above)
LAYER_IMPL_METRICS.record_redownloaded_after(since_last_eviction);
Ok((ResidentOrWantedEvicted::Resident(res), permit))
}
let res = Arc::new(DownloadedLayer {
owner: Arc::downgrade(self),
kind: tokio::sync::OnceCell::default(),
version: next_version,
});
self.access_stats.record_residence_event(
LayerResidenceStatus::Resident,
LayerResidenceEventReason::ResidenceChange,
);
let waiters = self.inner.initializer_count();
if waiters > 0 {
tracing::info!(waiters, "completing the on-demand download for other tasks");
}
scopeguard::ScopeGuard::into_inner(init_cancelled);
Ok((ResidentOrWantedEvicted::Resident(res), permit))
.instrument(tracing::info_span!("get_or_maybe_download", layer=%self))
};
if let Some(init_permit) = init_permit.take() {

View File

@@ -138,7 +138,7 @@ pub(super) async fn connection_manager_loop_step(
Ok(Some(broker_update)) => connection_manager_state.register_timeline_update(broker_update),
Err(status) => {
match status.code() {
Code::Unknown if status.message().contains("stream closed because of a broken pipe") => {
Code::Unknown if status.message().contains("stream closed because of a broken pipe") || status.message().contains("connection reset") => {
// tonic's error handling doesn't provide a clear code for disconnections: we get
// "h2 protocol error: error reading a body from connection: stream closed because of a broken pipe"
info!("broker disconnected: {status}");

View File

@@ -19,20 +19,21 @@
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include <curl/curl.h>
#include "access/xact.h"
#include "commands/defrem.h"
#include "fmgr.h"
#include "libpq/crypt.h"
#include "miscadmin.h"
#include "tcop/pquery.h"
#include "tcop/utility.h"
#include "access/xact.h"
#include "utils/acl.h"
#include "utils/guc.h"
#include "utils/hsearch.h"
#include "utils/memutils.h"
#include "commands/defrem.h"
#include "miscadmin.h"
#include "utils/acl.h"
#include "fmgr.h"
#include "utils/guc.h"
#include "port.h"
#include <curl/curl.h>
#include "utils/jsonb.h"
#include "libpq/crypt.h"
static ProcessUtility_hook_type PreviousProcessUtilityHook = NULL;

View File

@@ -1,4 +1,3 @@
/*-------------------------------------------------------------------------
*
* extension_server.c
@@ -10,21 +9,11 @@
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "tcop/pquery.h"
#include "tcop/utility.h"
#include "access/xact.h"
#include "utils/hsearch.h"
#include "utils/memutils.h"
#include "commands/defrem.h"
#include "miscadmin.h"
#include "utils/acl.h"
#include "fmgr.h"
#include "utils/guc.h"
#include "port.h"
#include "fmgr.h"
#include <curl/curl.h>
#include "utils/guc.h"
static int extension_server_port = 0;
static download_extension_file_hook_type prev_download_extension_file_hook = NULL;

View File

@@ -13,32 +13,30 @@
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include <sys/file.h>
#include <unistd.h>
#include <fcntl.h>
#include "postgres.h"
#include "neon_pgversioncompat.h"
#include "access/parallel.h"
#include "funcapi.h"
#include "miscadmin.h"
#include "pgstat.h"
#include "pagestore_client.h"
#include "access/parallel.h"
#include "pgstat.h"
#include "postmaster/bgworker.h"
#include RELFILEINFO_HDR
#include "storage/buf_internals.h"
#include "storage/latch.h"
#include "storage/fd.h"
#include "storage/ipc.h"
#include "storage/latch.h"
#include "storage/lwlock.h"
#include "storage/pg_shmem.h"
#include "utils/builtins.h"
#include "utils/dynahash.h"
#include "utils/guc.h"
#include "storage/fd.h"
#include "storage/pg_shmem.h"
#include "storage/buf_internals.h"
#include "pgstat.h"
/*
* Local file cache is used to temporary store relations pages in local file system.
@@ -102,8 +100,6 @@ static shmem_request_hook_type prev_shmem_request_hook;
#define LFC_ENABLED() (lfc_ctl->limit != 0)
void PGDLLEXPORT FileCacheMonitorMain(Datum main_arg);
/*
* Local file cache is optional and Neon can work without it.
* In case of any any errors with this cache, we should disable it but to not throw error.

View File

@@ -14,28 +14,24 @@
*/
#include "postgres.h"
#include "pagestore_client.h"
#include "fmgr.h"
#include "access/xlog.h"
#include "access/xlogutils.h"
#include "storage/buf_internals.h"
#include "storage/lwlock.h"
#include "storage/ipc.h"
#include "storage/pg_shmem.h"
#include "c.h"
#include "postmaster/interrupt.h"
#include "fmgr.h"
#include "libpq-fe.h"
#include "libpq/pqformat.h"
#include "libpq/libpq.h"
#include "libpq/pqformat.h"
#include "miscadmin.h"
#include "pgstat.h"
#include "postmaster/interrupt.h"
#include "storage/buf_internals.h"
#include "storage/ipc.h"
#include "storage/lwlock.h"
#include "storage/pg_shmem.h"
#include "utils/guc.h"
#include "neon.h"
#include "walproposer.h"
#include "neon_utils.h"
#include "pagestore_client.h"
#include "walproposer.h"
#define PageStoreTrace DEBUG5
@@ -62,8 +58,8 @@ char *neon_auth_token;
int readahead_buffer_size = 128;
int flush_every_n_requests = 8;
int n_reconnect_attempts = 0;
int max_reconnect_attempts = 60;
static int n_reconnect_attempts = 0;
static int max_reconnect_attempts = 60;
#define MAX_PAGESERVER_CONNSTRING_SIZE 256
@@ -83,8 +79,6 @@ static PagestoreShmemState *pagestore_shared;
static uint64 pagestore_local_counter = 0;
static char local_pageserver_connstring[MAX_PAGESERVER_CONNSTRING_SIZE];
bool (*old_redo_read_buffer_filter) (XLogReaderState *record, uint8 block_id) = NULL;
static bool pageserver_flush(void);
static void pageserver_disconnect(void);
@@ -627,8 +621,6 @@ pg_init_libpagestore(void)
smgr_hook = smgr_neon;
smgr_init_hook = smgr_init_neon;
dbsize_hook = neon_dbsize;
old_redo_read_buffer_filter = redo_read_buffer_filter;
redo_read_buffer_filter = neon_redo_read_buffer_filter;
}
lfc_init();

View File

@@ -27,13 +27,6 @@ extern void pg_init_walproposer(void);
extern void pg_init_extension_server(void);
/*
* Returns true if we shouldn't do REDO on that block in record indicated by
* block_id; false otherwise.
*/
extern bool neon_redo_read_buffer_filter(XLogReaderState *record, uint8 block_id);
extern bool (*old_redo_read_buffer_filter) (XLogReaderState *record, uint8 block_id);
extern uint64 BackpressureThrottlingTime(void);
extern void replication_feedback_get_lsns(XLogRecPtr *writeLsn, XLogRecPtr *flushLsn, XLogRecPtr *applyLsn);

View File

@@ -3,33 +3,8 @@
#include "postgres.h"
#include "access/timeline.h"
#include "access/xlogutils.h"
#include "common/logging.h"
#include "common/ip.h"
#include "funcapi.h"
#include "libpq/libpq.h"
#include "lib/stringinfo.h"
#include "libpq/pqformat.h"
#include "miscadmin.h"
#include "postmaster/interrupt.h"
#include "replication/slot.h"
#include "replication/walsender_private.h"
#include "storage/ipc.h"
#include "utils/builtins.h"
#include "utils/ps_status.h"
#include "libpq-fe.h"
#include <netinet/tcp.h>
#include <unistd.h>
#if PG_VERSION_NUM >= 150000
#include "access/xlogutils.h"
#include "access/xlogrecovery.h"
#endif
#if PG_MAJORVERSION_NUM >= 16
#include "utils/guc.h"
#endif
/*
* Convert a character which represents a hexadecimal digit to an integer.

View File

@@ -1,8 +1,6 @@
#ifndef __NEON_UTILS_H__
#define __NEON_UTILS_H__
#include "postgres.h"
bool HexDecodeString(uint8 *result, char *input, int nbytes);
uint32 pq_getmsgint32_le(StringInfo msg);
uint64 pq_getmsgint64_le(StringInfo msg);

View File

@@ -13,19 +13,16 @@
#ifndef pageserver_h
#define pageserver_h
#include "postgres.h"
#include "neon_pgversioncompat.h"
#include "access/xlogdefs.h"
#include RELFILEINFO_HDR
#include "storage/block.h"
#include "storage/smgr.h"
#include "lib/stringinfo.h"
#include "libpq/pqformat.h"
#include "storage/block.h"
#include "storage/smgr.h"
#include "utils/memutils.h"
#include "pg_config.h"
typedef enum
{
/* pagestore_client -> pagestore */
@@ -158,11 +155,8 @@ extern page_server_api *page_server;
extern char *page_server_connstring;
extern int flush_every_n_requests;
extern int readahead_buffer_size;
extern bool seqscan_prefetch_enabled;
extern int seqscan_prefetch_distance;
extern char *neon_timeline;
extern char *neon_tenant;
extern bool wal_redo;
extern int32 max_cluster_size;
extern const f_smgr *smgr_neon(BackendId backend, NRelFileInfo rinfo);

View File

@@ -47,25 +47,26 @@
#include "access/xact.h"
#include "access/xlog.h"
#include "access/xlogdefs.h"
#include "access/xloginsert.h"
#include "access/xlog_internal.h"
#include "access/xlogdefs.h"
#include "access/xlogutils.h"
#include "catalog/pg_class.h"
#include "common/hashfn.h"
#include "executor/instrument.h"
#include "pagestore_client.h"
#include "postmaster/interrupt.h"
#include "pgstat.h"
#include "postmaster/autovacuum.h"
#include "postmaster/interrupt.h"
#include "replication/walsender.h"
#include "storage/bufmgr.h"
#include "storage/buf_internals.h"
#include "storage/fsm_internals.h"
#include "storage/smgr.h"
#include "storage/md.h"
#include "pgstat.h"
#include "storage/smgr.h"
#include "pagestore_client.h"
#if PG_VERSION_NUM >= 150000
#include "access/xlogutils.h"
#include "access/xlogrecovery.h"
#endif
@@ -106,6 +107,9 @@ typedef enum
static SMgrRelation unlogged_build_rel = NULL;
static UnloggedBuildPhase unlogged_build_phase = UNLOGGED_BUILD_NOT_IN_PROGRESS;
static bool neon_redo_read_buffer_filter(XLogReaderState *record, uint8 block_id);
static bool (*old_redo_read_buffer_filter) (XLogReaderState *record, uint8 block_id) = NULL;
/*
* Prefetch implementation:
*
@@ -239,7 +243,7 @@ typedef struct PrefetchState
PrefetchRequest prf_buffer[]; /* prefetch buffers */
} PrefetchState;
PrefetchState *MyPState;
static PrefetchState *MyPState;
#define GetPrfSlot(ring_index) ( \
( \
@@ -257,7 +261,7 @@ PrefetchState *MyPState;
) \
)
XLogRecPtr prefetch_lsn = 0;
static XLogRecPtr prefetch_lsn = 0;
static bool compact_prefetch_buffers(void);
static void consume_prefetch_responses(void);
@@ -1371,6 +1375,9 @@ neon_init(void)
MyPState->prf_hash = prfh_create(MyPState->hashctx,
readahead_buffer_size, NULL);
old_redo_read_buffer_filter = redo_read_buffer_filter;
redo_read_buffer_filter = neon_redo_read_buffer_filter;
#ifdef DEBUG_COMPARE_LOCAL
mdinit();
#endif
@@ -2869,7 +2876,7 @@ get_fsm_physical_block(BlockNumber heapblk)
* contents, where with REDO locking it would wait on block 1 and see
* block 3 with post-REDO contents only.
*/
bool
static bool
neon_redo_read_buffer_filter(XLogReaderState *record, uint8 block_id)
{
XLogRecPtr end_recptr = record->EndRecPtr;

View File

@@ -1,14 +1,12 @@
#ifndef __NEON_WALPROPOSER_H__
#define __NEON_WALPROPOSER_H__
#include "postgres.h"
#include "access/xlogdefs.h"
#include "port.h"
#include "access/xlog_internal.h"
#include "access/transam.h"
#include "access/xlogdefs.h"
#include "access/xlog_internal.h"
#include "nodes/replnodes.h"
#include "utils/uuid.h"
#include "replication/walreceiver.h"
#include "utils/uuid.h"
#define SK_MAGIC 0xCafeCeefu
#define SK_PROTOCOL_VERSION 2

View File

@@ -3,11 +3,13 @@
* This is needed to avoid linking to full postgres server installation. This file
* is compiled as a part of libwalproposer static library.
*/
#include "postgres.h"
#include <stdio.h>
#include "walproposer.h"
#include "utils/datetime.h"
#include "miscadmin.h"
#include "utils/datetime.h"
#include "walproposer.h"
void
ExceptionalCondition(const char *conditionName,

View File

@@ -1482,6 +1482,21 @@ walprop_pg_wait_event_set(WalProposer *wp, long timeout, Safekeeper **sk, uint32
#if PG_MAJORVERSION_NUM >= 16
if (WalSndCtl != NULL)
ConditionVariablePrepareToSleep(&WalSndCtl->wal_flush_cv);
/*
* Now that we prepared the condvar, check flush ptr again -- it might have
* changed before we subscribed to cv so we missed the wakeup.
*
* Do that only when we're interested in new WAL: without sync-safekeepers
* and if election already passed.
*/
if (!wp->config->syncSafekeepers && wp->availableLsn != InvalidXLogRecPtr && GetFlushRecPtr(NULL) > wp->availableLsn)
{
ConditionVariableCancelSleep();
ResetLatch(MyLatch);
*events = WL_LATCH_SET;
return 1;
}
#endif
/*

258
poetry.lock generated
View File

@@ -2092,51 +2092,61 @@ files = [
[[package]]
name = "pyyaml"
version = "6.0"
version = "6.0.1"
description = "YAML parser and emitter for Python"
optional = false
python-versions = ">=3.6"
files = [
{file = "PyYAML-6.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:d4db7c7aef085872ef65a8fd7d6d09a14ae91f691dec3e87ee5ee0539d516f53"},
{file = "PyYAML-6.0-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:9df7ed3b3d2e0ecfe09e14741b857df43adb5a3ddadc919a2d94fbdf78fea53c"},
{file = "PyYAML-6.0-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:77f396e6ef4c73fdc33a9157446466f1cff553d979bd00ecb64385760c6babdc"},
{file = "PyYAML-6.0-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:a80a78046a72361de73f8f395f1f1e49f956c6be882eed58505a15f3e430962b"},
{file = "PyYAML-6.0-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl", hash = "sha256:f84fbc98b019fef2ee9a1cb3ce93e3187a6df0b2538a651bfb890254ba9f90b5"},
{file = "PyYAML-6.0-cp310-cp310-win32.whl", hash = "sha256:2cd5df3de48857ed0544b34e2d40e9fac445930039f3cfe4bcc592a1f836d513"},
{file = "PyYAML-6.0-cp310-cp310-win_amd64.whl", hash = "sha256:daf496c58a8c52083df09b80c860005194014c3698698d1a57cbcfa182142a3a"},
{file = "PyYAML-6.0-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:d4b0ba9512519522b118090257be113b9468d804b19d63c71dbcf4a48fa32358"},
{file = "PyYAML-6.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:81957921f441d50af23654aa6c5e5eaf9b06aba7f0a19c18a538dc7ef291c5a1"},
{file = "PyYAML-6.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:afa17f5bc4d1b10afd4466fd3a44dc0e245382deca5b3c353d8b757f9e3ecb8d"},
{file = "PyYAML-6.0-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:dbad0e9d368bb989f4515da330b88a057617d16b6a8245084f1b05400f24609f"},
{file = "PyYAML-6.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:432557aa2c09802be39460360ddffd48156e30721f5e8d917f01d31694216782"},
{file = "PyYAML-6.0-cp311-cp311-win32.whl", hash = "sha256:bfaef573a63ba8923503d27530362590ff4f576c626d86a9fed95822a8255fd7"},
{file = "PyYAML-6.0-cp311-cp311-win_amd64.whl", hash = "sha256:01b45c0191e6d66c470b6cf1b9531a771a83c1c4208272ead47a3ae4f2f603bf"},
{file = "PyYAML-6.0-cp36-cp36m-macosx_10_9_x86_64.whl", hash = "sha256:897b80890765f037df3403d22bab41627ca8811ae55e9a722fd0392850ec4d86"},
{file = "PyYAML-6.0-cp36-cp36m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:50602afada6d6cbfad699b0c7bb50d5ccffa7e46a3d738092afddc1f9758427f"},
{file = "PyYAML-6.0-cp36-cp36m-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:48c346915c114f5fdb3ead70312bd042a953a8ce5c7106d5bfb1a5254e47da92"},
{file = "PyYAML-6.0-cp36-cp36m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl", hash = "sha256:98c4d36e99714e55cfbaaee6dd5badbc9a1ec339ebfc3b1f52e293aee6bb71a4"},
{file = "PyYAML-6.0-cp36-cp36m-win32.whl", hash = "sha256:0283c35a6a9fbf047493e3a0ce8d79ef5030852c51e9d911a27badfde0605293"},
{file = "PyYAML-6.0-cp36-cp36m-win_amd64.whl", hash = "sha256:07751360502caac1c067a8132d150cf3d61339af5691fe9e87803040dbc5db57"},
{file = "PyYAML-6.0-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:819b3830a1543db06c4d4b865e70ded25be52a2e0631ccd2f6a47a2822f2fd7c"},
{file = "PyYAML-6.0-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:473f9edb243cb1935ab5a084eb238d842fb8f404ed2193a915d1784b5a6b5fc0"},
{file = "PyYAML-6.0-cp37-cp37m-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:0ce82d761c532fe4ec3f87fc45688bdd3a4c1dc5e0b4a19814b9009a29baefd4"},
{file = "PyYAML-6.0-cp37-cp37m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl", hash = "sha256:231710d57adfd809ef5d34183b8ed1eeae3f76459c18fb4a0b373ad56bedcdd9"},
{file = "PyYAML-6.0-cp37-cp37m-win32.whl", hash = "sha256:c5687b8d43cf58545ade1fe3e055f70eac7a5a1a0bf42824308d868289a95737"},
{file = "PyYAML-6.0-cp37-cp37m-win_amd64.whl", hash = "sha256:d15a181d1ecd0d4270dc32edb46f7cb7733c7c508857278d3d378d14d606db2d"},
{file = "PyYAML-6.0-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:0b4624f379dab24d3725ffde76559cff63d9ec94e1736b556dacdfebe5ab6d4b"},
{file = "PyYAML-6.0-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:213c60cd50106436cc818accf5baa1aba61c0189ff610f64f4a3e8c6726218ba"},
{file = "PyYAML-6.0-cp38-cp38-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:9fa600030013c4de8165339db93d182b9431076eb98eb40ee068700c9c813e34"},
{file = "PyYAML-6.0-cp38-cp38-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl", hash = "sha256:277a0ef2981ca40581a47093e9e2d13b3f1fbbeffae064c1d21bfceba2030287"},
{file = "PyYAML-6.0-cp38-cp38-win32.whl", hash = "sha256:d4eccecf9adf6fbcc6861a38015c2a64f38b9d94838ac1810a9023a0609e1b78"},
{file = "PyYAML-6.0-cp38-cp38-win_amd64.whl", hash = "sha256:1e4747bc279b4f613a09eb64bba2ba602d8a6664c6ce6396a4d0cd413a50ce07"},
{file = "PyYAML-6.0-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:055d937d65826939cb044fc8c9b08889e8c743fdc6a32b33e2390f66013e449b"},
{file = "PyYAML-6.0-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:e61ceaab6f49fb8bdfaa0f92c4b57bcfbea54c09277b1b4f7ac376bfb7a7c174"},
{file = "PyYAML-6.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d67d839ede4ed1b28a4e8909735fc992a923cdb84e618544973d7dfc71540803"},
{file = "PyYAML-6.0-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:cba8c411ef271aa037d7357a2bc8f9ee8b58b9965831d9e51baf703280dc73d3"},
{file = "PyYAML-6.0-cp39-cp39-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl", hash = "sha256:40527857252b61eacd1d9af500c3337ba8deb8fc298940291486c465c8b46ec0"},
{file = "PyYAML-6.0-cp39-cp39-win32.whl", hash = "sha256:b5b9eccad747aabaaffbc6064800670f0c297e52c12754eb1d976c57e4f74dcb"},
{file = "PyYAML-6.0-cp39-cp39-win_amd64.whl", hash = "sha256:b3d267842bf12586ba6c734f89d1f5b871df0273157918b0ccefa29deb05c21c"},
{file = "PyYAML-6.0.tar.gz", hash = "sha256:68fb519c14306fec9720a2a5b45bc9f0c8d1b9c72adf45c37baedfcd949c35a2"},
{file = "PyYAML-6.0.1-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:d858aa552c999bc8a8d57426ed01e40bef403cd8ccdd0fc5f6f04a00414cac2a"},
{file = "PyYAML-6.0.1-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:fd66fc5d0da6d9815ba2cebeb4205f95818ff4b79c3ebe268e75d961704af52f"},
{file = "PyYAML-6.0.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:69b023b2b4daa7548bcfbd4aa3da05b3a74b772db9e23b982788168117739938"},
{file = "PyYAML-6.0.1-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:81e0b275a9ecc9c0c0c07b4b90ba548307583c125f54d5b6946cfee6360c733d"},
{file = "PyYAML-6.0.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:ba336e390cd8e4d1739f42dfe9bb83a3cc2e80f567d8805e11b46f4a943f5515"},
{file = "PyYAML-6.0.1-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:326c013efe8048858a6d312ddd31d56e468118ad4cdeda36c719bf5bb6192290"},
{file = "PyYAML-6.0.1-cp310-cp310-win32.whl", hash = "sha256:bd4af7373a854424dabd882decdc5579653d7868b8fb26dc7d0e99f823aa5924"},
{file = "PyYAML-6.0.1-cp310-cp310-win_amd64.whl", hash = "sha256:fd1592b3fdf65fff2ad0004b5e363300ef59ced41c2e6b3a99d4089fa8c5435d"},
{file = "PyYAML-6.0.1-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:6965a7bc3cf88e5a1c3bd2e0b5c22f8d677dc88a455344035f03399034eb3007"},
{file = "PyYAML-6.0.1-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:f003ed9ad21d6a4713f0a9b5a7a0a79e08dd0f221aff4525a2be4c346ee60aab"},
{file = "PyYAML-6.0.1-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:42f8152b8dbc4fe7d96729ec2b99c7097d656dc1213a3229ca5383f973a5ed6d"},
{file = "PyYAML-6.0.1-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:062582fca9fabdd2c8b54a3ef1c978d786e0f6b3a1510e0ac93ef59e0ddae2bc"},
{file = "PyYAML-6.0.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:d2b04aac4d386b172d5b9692e2d2da8de7bfb6c387fa4f801fbf6fb2e6ba4673"},
{file = "PyYAML-6.0.1-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:e7d73685e87afe9f3b36c799222440d6cf362062f78be1013661b00c5c6f678b"},
{file = "PyYAML-6.0.1-cp311-cp311-win32.whl", hash = "sha256:1635fd110e8d85d55237ab316b5b011de701ea0f29d07611174a1b42f1444741"},
{file = "PyYAML-6.0.1-cp311-cp311-win_amd64.whl", hash = "sha256:bf07ee2fef7014951eeb99f56f39c9bb4af143d8aa3c21b1677805985307da34"},
{file = "PyYAML-6.0.1-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:855fb52b0dc35af121542a76b9a84f8d1cd886ea97c84703eaa6d88e37a2ad28"},
{file = "PyYAML-6.0.1-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:40df9b996c2b73138957fe23a16a4f0ba614f4c0efce1e9406a184b6d07fa3a9"},
{file = "PyYAML-6.0.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:6c22bec3fbe2524cde73d7ada88f6566758a8f7227bfbf93a408a9d86bcc12a0"},
{file = "PyYAML-6.0.1-cp312-cp312-musllinux_1_1_x86_64.whl", hash = "sha256:8d4e9c88387b0f5c7d5f281e55304de64cf7f9c0021a3525bd3b1c542da3b0e4"},
{file = "PyYAML-6.0.1-cp312-cp312-win32.whl", hash = "sha256:d483d2cdf104e7c9fa60c544d92981f12ad66a457afae824d146093b8c294c54"},
{file = "PyYAML-6.0.1-cp312-cp312-win_amd64.whl", hash = "sha256:0d3304d8c0adc42be59c5f8a4d9e3d7379e6955ad754aa9d6ab7a398b59dd1df"},
{file = "PyYAML-6.0.1-cp36-cp36m-macosx_10_9_x86_64.whl", hash = "sha256:50550eb667afee136e9a77d6dc71ae76a44df8b3e51e41b77f6de2932bfe0f47"},
{file = "PyYAML-6.0.1-cp36-cp36m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:1fe35611261b29bd1de0070f0b2f47cb6ff71fa6595c077e42bd0c419fa27b98"},
{file = "PyYAML-6.0.1-cp36-cp36m-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:704219a11b772aea0d8ecd7058d0082713c3562b4e271b849ad7dc4a5c90c13c"},
{file = "PyYAML-6.0.1-cp36-cp36m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:afd7e57eddb1a54f0f1a974bc4391af8bcce0b444685d936840f125cf046d5bd"},
{file = "PyYAML-6.0.1-cp36-cp36m-win32.whl", hash = "sha256:fca0e3a251908a499833aa292323f32437106001d436eca0e6e7833256674585"},
{file = "PyYAML-6.0.1-cp36-cp36m-win_amd64.whl", hash = "sha256:f22ac1c3cac4dbc50079e965eba2c1058622631e526bd9afd45fedd49ba781fa"},
{file = "PyYAML-6.0.1-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:b1275ad35a5d18c62a7220633c913e1b42d44b46ee12554e5fd39c70a243d6a3"},
{file = "PyYAML-6.0.1-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:18aeb1bf9a78867dc38b259769503436b7c72f7a1f1f4c93ff9a17de54319b27"},
{file = "PyYAML-6.0.1-cp37-cp37m-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:596106435fa6ad000c2991a98fa58eeb8656ef2325d7e158344fb33864ed87e3"},
{file = "PyYAML-6.0.1-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:baa90d3f661d43131ca170712d903e6295d1f7a0f595074f151c0aed377c9b9c"},
{file = "PyYAML-6.0.1-cp37-cp37m-win32.whl", hash = "sha256:9046c58c4395dff28dd494285c82ba00b546adfc7ef001486fbf0324bc174fba"},
{file = "PyYAML-6.0.1-cp37-cp37m-win_amd64.whl", hash = "sha256:4fb147e7a67ef577a588a0e2c17b6db51dda102c71de36f8549b6816a96e1867"},
{file = "PyYAML-6.0.1-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:1d4c7e777c441b20e32f52bd377e0c409713e8bb1386e1099c2415f26e479595"},
{file = "PyYAML-6.0.1-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:a0cd17c15d3bb3fa06978b4e8958dcdc6e0174ccea823003a106c7d4d7899ac5"},
{file = "PyYAML-6.0.1-cp38-cp38-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:28c119d996beec18c05208a8bd78cbe4007878c6dd15091efb73a30e90539696"},
{file = "PyYAML-6.0.1-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:7e07cbde391ba96ab58e532ff4803f79c4129397514e1413a7dc761ccd755735"},
{file = "PyYAML-6.0.1-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:49a183be227561de579b4a36efbb21b3eab9651dd81b1858589f796549873dd6"},
{file = "PyYAML-6.0.1-cp38-cp38-win32.whl", hash = "sha256:184c5108a2aca3c5b3d3bf9395d50893a7ab82a38004c8f61c258d4428e80206"},
{file = "PyYAML-6.0.1-cp38-cp38-win_amd64.whl", hash = "sha256:1e2722cc9fbb45d9b87631ac70924c11d3a401b2d7f410cc0e3bbf249f2dca62"},
{file = "PyYAML-6.0.1-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:9eb6caa9a297fc2c2fb8862bc5370d0303ddba53ba97e71f08023b6cd73d16a8"},
{file = "PyYAML-6.0.1-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:c8098ddcc2a85b61647b2590f825f3db38891662cfc2fc776415143f599bb859"},
{file = "PyYAML-6.0.1-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:5773183b6446b2c99bb77e77595dd486303b4faab2b086e7b17bc6bef28865f6"},
{file = "PyYAML-6.0.1-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:b786eecbdf8499b9ca1d697215862083bd6d2a99965554781d0d8d1ad31e13a0"},
{file = "PyYAML-6.0.1-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:bc1bf2925a1ecd43da378f4db9e4f799775d6367bdb94671027b73b393a7c42c"},
{file = "PyYAML-6.0.1-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:04ac92ad1925b2cff1db0cfebffb6ffc43457495c9b3c39d3fcae417d7125dc5"},
{file = "PyYAML-6.0.1-cp39-cp39-win32.whl", hash = "sha256:faca3bdcf85b2fc05d06ff3fbc1f83e1391b3e724afa3feba7d13eeab355484c"},
{file = "PyYAML-6.0.1-cp39-cp39-win_amd64.whl", hash = "sha256:510c9deebc5c0225e8c96813043e62b680ba2f9c50a08d3724c7f28a747d1486"},
{file = "PyYAML-6.0.1.tar.gz", hash = "sha256:bfdf460b1736c775f2ba9f6a92bca30bc2095067b8a9d77876d1fad6cc3b4a43"},
]
[[package]]
@@ -2553,85 +2563,101 @@ files = [
[[package]]
name = "yarl"
version = "1.8.2"
version = "1.9.4"
description = "Yet another URL library"
optional = false
python-versions = ">=3.7"
files = [
{file = "yarl-1.8.2-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:bb81f753c815f6b8e2ddd2eef3c855cf7da193b82396ac013c661aaa6cc6b0a5"},
{file = "yarl-1.8.2-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:47d49ac96156f0928f002e2424299b2c91d9db73e08c4cd6742923a086f1c863"},
{file = "yarl-1.8.2-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:3fc056e35fa6fba63248d93ff6e672c096f95f7836938241ebc8260e062832fe"},
{file = "yarl-1.8.2-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:58a3c13d1c3005dbbac5c9f0d3210b60220a65a999b1833aa46bd6677c69b08e"},
{file = "yarl-1.8.2-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:10b08293cda921157f1e7c2790999d903b3fd28cd5c208cf8826b3b508026996"},
{file = "yarl-1.8.2-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:de986979bbd87272fe557e0a8fcb66fd40ae2ddfe28a8b1ce4eae22681728fef"},
{file = "yarl-1.8.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:6c4fcfa71e2c6a3cb568cf81aadc12768b9995323186a10827beccf5fa23d4f8"},
{file = "yarl-1.8.2-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:ae4d7ff1049f36accde9e1ef7301912a751e5bae0a9d142459646114c70ecba6"},
{file = "yarl-1.8.2-cp310-cp310-musllinux_1_1_aarch64.whl", hash = "sha256:bf071f797aec5b96abfc735ab97da9fd8f8768b43ce2abd85356a3127909d146"},
{file = "yarl-1.8.2-cp310-cp310-musllinux_1_1_i686.whl", hash = "sha256:74dece2bfc60f0f70907c34b857ee98f2c6dd0f75185db133770cd67300d505f"},
{file = "yarl-1.8.2-cp310-cp310-musllinux_1_1_ppc64le.whl", hash = "sha256:df60a94d332158b444301c7f569659c926168e4d4aad2cfbf4bce0e8fb8be826"},
{file = "yarl-1.8.2-cp310-cp310-musllinux_1_1_s390x.whl", hash = "sha256:63243b21c6e28ec2375f932a10ce7eda65139b5b854c0f6b82ed945ba526bff3"},
{file = "yarl-1.8.2-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:cfa2bbca929aa742b5084fd4663dd4b87c191c844326fcb21c3afd2d11497f80"},
{file = "yarl-1.8.2-cp310-cp310-win32.whl", hash = "sha256:b05df9ea7496df11b710081bd90ecc3a3db6adb4fee36f6a411e7bc91a18aa42"},
{file = "yarl-1.8.2-cp310-cp310-win_amd64.whl", hash = "sha256:24ad1d10c9db1953291f56b5fe76203977f1ed05f82d09ec97acb623a7976574"},
{file = "yarl-1.8.2-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:2a1fca9588f360036242f379bfea2b8b44cae2721859b1c56d033adfd5893634"},
{file = "yarl-1.8.2-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:f37db05c6051eff17bc832914fe46869f8849de5b92dc4a3466cd63095d23dfd"},
{file = "yarl-1.8.2-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:77e913b846a6b9c5f767b14dc1e759e5aff05502fe73079f6f4176359d832581"},
{file = "yarl-1.8.2-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:0978f29222e649c351b173da2b9b4665ad1feb8d1daa9d971eb90df08702668a"},
{file = "yarl-1.8.2-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:388a45dc77198b2460eac0aca1efd6a7c09e976ee768b0d5109173e521a19daf"},
{file = "yarl-1.8.2-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:2305517e332a862ef75be8fad3606ea10108662bc6fe08509d5ca99503ac2aee"},
{file = "yarl-1.8.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:42430ff511571940d51e75cf42f1e4dbdded477e71c1b7a17f4da76c1da8ea76"},
{file = "yarl-1.8.2-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:3150078118f62371375e1e69b13b48288e44f6691c1069340081c3fd12c94d5b"},
{file = "yarl-1.8.2-cp311-cp311-musllinux_1_1_aarch64.whl", hash = "sha256:c15163b6125db87c8f53c98baa5e785782078fbd2dbeaa04c6141935eb6dab7a"},
{file = "yarl-1.8.2-cp311-cp311-musllinux_1_1_i686.whl", hash = "sha256:4d04acba75c72e6eb90745447d69f84e6c9056390f7a9724605ca9c56b4afcc6"},
{file = "yarl-1.8.2-cp311-cp311-musllinux_1_1_ppc64le.whl", hash = "sha256:e7fd20d6576c10306dea2d6a5765f46f0ac5d6f53436217913e952d19237efc4"},
{file = "yarl-1.8.2-cp311-cp311-musllinux_1_1_s390x.whl", hash = "sha256:75c16b2a900b3536dfc7014905a128a2bea8fb01f9ee26d2d7d8db0a08e7cb2c"},
{file = "yarl-1.8.2-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:6d88056a04860a98341a0cf53e950e3ac9f4e51d1b6f61a53b0609df342cc8b2"},
{file = "yarl-1.8.2-cp311-cp311-win32.whl", hash = "sha256:fb742dcdd5eec9f26b61224c23baea46c9055cf16f62475e11b9b15dfd5c117b"},
{file = "yarl-1.8.2-cp311-cp311-win_amd64.whl", hash = "sha256:8c46d3d89902c393a1d1e243ac847e0442d0196bbd81aecc94fcebbc2fd5857c"},
{file = "yarl-1.8.2-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:ceff9722e0df2e0a9e8a79c610842004fa54e5b309fe6d218e47cd52f791d7ef"},
{file = "yarl-1.8.2-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:3f6b4aca43b602ba0f1459de647af954769919c4714706be36af670a5f44c9c1"},
{file = "yarl-1.8.2-cp37-cp37m-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:1684a9bd9077e922300ecd48003ddae7a7474e0412bea38d4631443a91d61077"},
{file = "yarl-1.8.2-cp37-cp37m-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:ebb78745273e51b9832ef90c0898501006670d6e059f2cdb0e999494eb1450c2"},
{file = "yarl-1.8.2-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:3adeef150d528ded2a8e734ebf9ae2e658f4c49bf413f5f157a470e17a4a2e89"},
{file = "yarl-1.8.2-cp37-cp37m-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:57a7c87927a468e5a1dc60c17caf9597161d66457a34273ab1760219953f7f4c"},
{file = "yarl-1.8.2-cp37-cp37m-musllinux_1_1_aarch64.whl", hash = "sha256:efff27bd8cbe1f9bd127e7894942ccc20c857aa8b5a0327874f30201e5ce83d0"},
{file = "yarl-1.8.2-cp37-cp37m-musllinux_1_1_i686.whl", hash = "sha256:a783cd344113cb88c5ff7ca32f1f16532a6f2142185147822187913eb989f739"},
{file = "yarl-1.8.2-cp37-cp37m-musllinux_1_1_ppc64le.whl", hash = "sha256:705227dccbe96ab02c7cb2c43e1228e2826e7ead880bb19ec94ef279e9555b5b"},
{file = "yarl-1.8.2-cp37-cp37m-musllinux_1_1_s390x.whl", hash = "sha256:34c09b43bd538bf6c4b891ecce94b6fa4f1f10663a8d4ca589a079a5018f6ed7"},
{file = "yarl-1.8.2-cp37-cp37m-musllinux_1_1_x86_64.whl", hash = "sha256:a48f4f7fea9a51098b02209d90297ac324241bf37ff6be6d2b0149ab2bd51b37"},
{file = "yarl-1.8.2-cp37-cp37m-win32.whl", hash = "sha256:0414fd91ce0b763d4eadb4456795b307a71524dbacd015c657bb2a39db2eab89"},
{file = "yarl-1.8.2-cp37-cp37m-win_amd64.whl", hash = "sha256:d881d152ae0007809c2c02e22aa534e702f12071e6b285e90945aa3c376463c5"},
{file = "yarl-1.8.2-cp38-cp38-macosx_10_9_universal2.whl", hash = "sha256:5df5e3d04101c1e5c3b1d69710b0574171cc02fddc4b23d1b2813e75f35a30b1"},
{file = "yarl-1.8.2-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:7a66c506ec67eb3159eea5096acd05f5e788ceec7b96087d30c7d2865a243918"},
{file = "yarl-1.8.2-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:2b4fa2606adf392051d990c3b3877d768771adc3faf2e117b9de7eb977741229"},
{file = "yarl-1.8.2-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:1e21fb44e1eff06dd6ef971d4bdc611807d6bd3691223d9c01a18cec3677939e"},
{file = "yarl-1.8.2-cp38-cp38-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:93202666046d9edadfe9f2e7bf5e0782ea0d497b6d63da322e541665d65a044e"},
{file = "yarl-1.8.2-cp38-cp38-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:fc77086ce244453e074e445104f0ecb27530d6fd3a46698e33f6c38951d5a0f1"},
{file = "yarl-1.8.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:64dd68a92cab699a233641f5929a40f02a4ede8c009068ca8aa1fe87b8c20ae3"},
{file = "yarl-1.8.2-cp38-cp38-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:1b372aad2b5f81db66ee7ec085cbad72c4da660d994e8e590c997e9b01e44901"},
{file = "yarl-1.8.2-cp38-cp38-musllinux_1_1_aarch64.whl", hash = "sha256:e6f3515aafe0209dd17fb9bdd3b4e892963370b3de781f53e1746a521fb39fc0"},
{file = "yarl-1.8.2-cp38-cp38-musllinux_1_1_i686.whl", hash = "sha256:dfef7350ee369197106805e193d420b75467b6cceac646ea5ed3049fcc950a05"},
{file = "yarl-1.8.2-cp38-cp38-musllinux_1_1_ppc64le.whl", hash = "sha256:728be34f70a190566d20aa13dc1f01dc44b6aa74580e10a3fb159691bc76909d"},
{file = "yarl-1.8.2-cp38-cp38-musllinux_1_1_s390x.whl", hash = "sha256:ff205b58dc2929191f68162633d5e10e8044398d7a45265f90a0f1d51f85f72c"},
{file = "yarl-1.8.2-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:baf211dcad448a87a0d9047dc8282d7de59473ade7d7fdf22150b1d23859f946"},
{file = "yarl-1.8.2-cp38-cp38-win32.whl", hash = "sha256:272b4f1599f1b621bf2aabe4e5b54f39a933971f4e7c9aa311d6d7dc06965165"},
{file = "yarl-1.8.2-cp38-cp38-win_amd64.whl", hash = "sha256:326dd1d3caf910cd26a26ccbfb84c03b608ba32499b5d6eeb09252c920bcbe4f"},
{file = "yarl-1.8.2-cp39-cp39-macosx_10_9_universal2.whl", hash = "sha256:f8ca8ad414c85bbc50f49c0a106f951613dfa5f948ab69c10ce9b128d368baf8"},
{file = "yarl-1.8.2-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:418857f837347e8aaef682679f41e36c24250097f9e2f315d39bae3a99a34cbf"},
{file = "yarl-1.8.2-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:ae0eec05ab49e91a78700761777f284c2df119376e391db42c38ab46fd662b77"},
{file = "yarl-1.8.2-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:009a028127e0a1755c38b03244c0bea9d5565630db9c4cf9572496e947137a87"},
{file = "yarl-1.8.2-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:3edac5d74bb3209c418805bda77f973117836e1de7c000e9755e572c1f7850d0"},
{file = "yarl-1.8.2-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:da65c3f263729e47351261351b8679c6429151ef9649bba08ef2528ff2c423b2"},
{file = "yarl-1.8.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0ef8fb25e52663a1c85d608f6dd72e19bd390e2ecaf29c17fb08f730226e3a08"},
{file = "yarl-1.8.2-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:bcd7bb1e5c45274af9a1dd7494d3c52b2be5e6bd8d7e49c612705fd45420b12d"},
{file = "yarl-1.8.2-cp39-cp39-musllinux_1_1_aarch64.whl", hash = "sha256:44ceac0450e648de86da8e42674f9b7077d763ea80c8ceb9d1c3e41f0f0a9951"},
{file = "yarl-1.8.2-cp39-cp39-musllinux_1_1_i686.whl", hash = "sha256:97209cc91189b48e7cfe777237c04af8e7cc51eb369004e061809bcdf4e55220"},
{file = "yarl-1.8.2-cp39-cp39-musllinux_1_1_ppc64le.whl", hash = "sha256:48dd18adcf98ea9cd721a25313aef49d70d413a999d7d89df44f469edfb38a06"},
{file = "yarl-1.8.2-cp39-cp39-musllinux_1_1_s390x.whl", hash = "sha256:e59399dda559688461762800d7fb34d9e8a6a7444fd76ec33220a926c8be1516"},
{file = "yarl-1.8.2-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:d617c241c8c3ad5c4e78a08429fa49e4b04bedfc507b34b4d8dceb83b4af3588"},
{file = "yarl-1.8.2-cp39-cp39-win32.whl", hash = "sha256:cb6d48d80a41f68de41212f3dfd1a9d9898d7841c8f7ce6696cf2fd9cb57ef83"},
{file = "yarl-1.8.2-cp39-cp39-win_amd64.whl", hash = "sha256:6604711362f2dbf7160df21c416f81fac0de6dbcf0b5445a2ef25478ecc4c778"},
{file = "yarl-1.8.2.tar.gz", hash = "sha256:49d43402c6e3013ad0978602bf6bf5328535c48d192304b91b97a3c6790b1562"},
{file = "yarl-1.9.4-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:a8c1df72eb746f4136fe9a2e72b0c9dc1da1cbd23b5372f94b5820ff8ae30e0e"},
{file = "yarl-1.9.4-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:a3a6ed1d525bfb91b3fc9b690c5a21bb52de28c018530ad85093cc488bee2dd2"},
{file = "yarl-1.9.4-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:c38c9ddb6103ceae4e4498f9c08fac9b590c5c71b0370f98714768e22ac6fa66"},
{file = "yarl-1.9.4-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d9e09c9d74f4566e905a0b8fa668c58109f7624db96a2171f21747abc7524234"},
{file = "yarl-1.9.4-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:b8477c1ee4bd47c57d49621a062121c3023609f7a13b8a46953eb6c9716ca392"},
{file = "yarl-1.9.4-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:d5ff2c858f5f6a42c2a8e751100f237c5e869cbde669a724f2062d4c4ef93551"},
{file = "yarl-1.9.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:357495293086c5b6d34ca9616a43d329317feab7917518bc97a08f9e55648455"},
{file = "yarl-1.9.4-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:54525ae423d7b7a8ee81ba189f131054defdb122cde31ff17477951464c1691c"},
{file = "yarl-1.9.4-cp310-cp310-musllinux_1_1_aarch64.whl", hash = "sha256:801e9264d19643548651b9db361ce3287176671fb0117f96b5ac0ee1c3530d53"},
{file = "yarl-1.9.4-cp310-cp310-musllinux_1_1_i686.whl", hash = "sha256:e516dc8baf7b380e6c1c26792610230f37147bb754d6426462ab115a02944385"},
{file = "yarl-1.9.4-cp310-cp310-musllinux_1_1_ppc64le.whl", hash = "sha256:7d5aaac37d19b2904bb9dfe12cdb08c8443e7ba7d2852894ad448d4b8f442863"},
{file = "yarl-1.9.4-cp310-cp310-musllinux_1_1_s390x.whl", hash = "sha256:54beabb809ffcacbd9d28ac57b0db46e42a6e341a030293fb3185c409e626b8b"},
{file = "yarl-1.9.4-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:bac8d525a8dbc2a1507ec731d2867025d11ceadcb4dd421423a5d42c56818541"},
{file = "yarl-1.9.4-cp310-cp310-win32.whl", hash = "sha256:7855426dfbddac81896b6e533ebefc0af2f132d4a47340cee6d22cac7190022d"},
{file = "yarl-1.9.4-cp310-cp310-win_amd64.whl", hash = "sha256:848cd2a1df56ddbffeb375535fb62c9d1645dde33ca4d51341378b3f5954429b"},
{file = "yarl-1.9.4-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:35a2b9396879ce32754bd457d31a51ff0a9d426fd9e0e3c33394bf4b9036b099"},
{file = "yarl-1.9.4-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:4c7d56b293cc071e82532f70adcbd8b61909eec973ae9d2d1f9b233f3d943f2c"},
{file = "yarl-1.9.4-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:d8a1c6c0be645c745a081c192e747c5de06e944a0d21245f4cf7c05e457c36e0"},
{file = "yarl-1.9.4-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:4b3c1ffe10069f655ea2d731808e76e0f452fc6c749bea04781daf18e6039525"},
{file = "yarl-1.9.4-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:549d19c84c55d11687ddbd47eeb348a89df9cb30e1993f1b128f4685cd0ebbf8"},
{file = "yarl-1.9.4-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:a7409f968456111140c1c95301cadf071bd30a81cbd7ab829169fb9e3d72eae9"},
{file = "yarl-1.9.4-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:e23a6d84d9d1738dbc6e38167776107e63307dfc8ad108e580548d1f2c587f42"},
{file = "yarl-1.9.4-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:d8b889777de69897406c9fb0b76cdf2fd0f31267861ae7501d93003d55f54fbe"},
{file = "yarl-1.9.4-cp311-cp311-musllinux_1_1_aarch64.whl", hash = "sha256:03caa9507d3d3c83bca08650678e25364e1843b484f19986a527630ca376ecce"},
{file = "yarl-1.9.4-cp311-cp311-musllinux_1_1_i686.whl", hash = "sha256:4e9035df8d0880b2f1c7f5031f33f69e071dfe72ee9310cfc76f7b605958ceb9"},
{file = "yarl-1.9.4-cp311-cp311-musllinux_1_1_ppc64le.whl", hash = "sha256:c0ec0ed476f77db9fb29bca17f0a8fcc7bc97ad4c6c1d8959c507decb22e8572"},
{file = "yarl-1.9.4-cp311-cp311-musllinux_1_1_s390x.whl", hash = "sha256:ee04010f26d5102399bd17f8df8bc38dc7ccd7701dc77f4a68c5b8d733406958"},
{file = "yarl-1.9.4-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:49a180c2e0743d5d6e0b4d1a9e5f633c62eca3f8a86ba5dd3c471060e352ca98"},
{file = "yarl-1.9.4-cp311-cp311-win32.whl", hash = "sha256:81eb57278deb6098a5b62e88ad8281b2ba09f2f1147c4767522353eaa6260b31"},
{file = "yarl-1.9.4-cp311-cp311-win_amd64.whl", hash = "sha256:d1d2532b340b692880261c15aee4dc94dd22ca5d61b9db9a8a361953d36410b1"},
{file = "yarl-1.9.4-cp312-cp312-macosx_10_9_universal2.whl", hash = "sha256:0d2454f0aef65ea81037759be5ca9947539667eecebca092733b2eb43c965a81"},
{file = "yarl-1.9.4-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:44d8ffbb9c06e5a7f529f38f53eda23e50d1ed33c6c869e01481d3fafa6b8142"},
{file = "yarl-1.9.4-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:aaaea1e536f98754a6e5c56091baa1b6ce2f2700cc4a00b0d49eca8dea471074"},
{file = "yarl-1.9.4-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:3777ce5536d17989c91696db1d459574e9a9bd37660ea7ee4d3344579bb6f129"},
{file = "yarl-1.9.4-cp312-cp312-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:9fc5fc1eeb029757349ad26bbc5880557389a03fa6ada41703db5e068881e5f2"},
{file = "yarl-1.9.4-cp312-cp312-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:ea65804b5dc88dacd4a40279af0cdadcfe74b3e5b4c897aa0d81cf86927fee78"},
{file = "yarl-1.9.4-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:aa102d6d280a5455ad6a0f9e6d769989638718e938a6a0a2ff3f4a7ff8c62cc4"},
{file = "yarl-1.9.4-cp312-cp312-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:09efe4615ada057ba2d30df871d2f668af661e971dfeedf0c159927d48bbeff0"},
{file = "yarl-1.9.4-cp312-cp312-musllinux_1_1_aarch64.whl", hash = "sha256:008d3e808d03ef28542372d01057fd09168419cdc8f848efe2804f894ae03e51"},
{file = "yarl-1.9.4-cp312-cp312-musllinux_1_1_i686.whl", hash = "sha256:6f5cb257bc2ec58f437da2b37a8cd48f666db96d47b8a3115c29f316313654ff"},
{file = "yarl-1.9.4-cp312-cp312-musllinux_1_1_ppc64le.whl", hash = "sha256:992f18e0ea248ee03b5a6e8b3b4738850ae7dbb172cc41c966462801cbf62cf7"},
{file = "yarl-1.9.4-cp312-cp312-musllinux_1_1_s390x.whl", hash = "sha256:0e9d124c191d5b881060a9e5060627694c3bdd1fe24c5eecc8d5d7d0eb6faabc"},
{file = "yarl-1.9.4-cp312-cp312-musllinux_1_1_x86_64.whl", hash = "sha256:3986b6f41ad22988e53d5778f91855dc0399b043fc8946d4f2e68af22ee9ff10"},
{file = "yarl-1.9.4-cp312-cp312-win32.whl", hash = "sha256:4b21516d181cd77ebd06ce160ef8cc2a5e9ad35fb1c5930882baff5ac865eee7"},
{file = "yarl-1.9.4-cp312-cp312-win_amd64.whl", hash = "sha256:a9bd00dc3bc395a662900f33f74feb3e757429e545d831eef5bb280252631984"},
{file = "yarl-1.9.4-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:63b20738b5aac74e239622d2fe30df4fca4942a86e31bf47a81a0e94c14df94f"},
{file = "yarl-1.9.4-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d7d7f7de27b8944f1fee2c26a88b4dabc2409d2fea7a9ed3df79b67277644e17"},
{file = "yarl-1.9.4-cp37-cp37m-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:c74018551e31269d56fab81a728f683667e7c28c04e807ba08f8c9e3bba32f14"},
{file = "yarl-1.9.4-cp37-cp37m-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:ca06675212f94e7a610e85ca36948bb8fc023e458dd6c63ef71abfd482481aa5"},
{file = "yarl-1.9.4-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:5aef935237d60a51a62b86249839b51345f47564208c6ee615ed2a40878dccdd"},
{file = "yarl-1.9.4-cp37-cp37m-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:2b134fd795e2322b7684155b7855cc99409d10b2e408056db2b93b51a52accc7"},
{file = "yarl-1.9.4-cp37-cp37m-musllinux_1_1_aarch64.whl", hash = "sha256:d25039a474c4c72a5ad4b52495056f843a7ff07b632c1b92ea9043a3d9950f6e"},
{file = "yarl-1.9.4-cp37-cp37m-musllinux_1_1_i686.whl", hash = "sha256:f7d6b36dd2e029b6bcb8a13cf19664c7b8e19ab3a58e0fefbb5b8461447ed5ec"},
{file = "yarl-1.9.4-cp37-cp37m-musllinux_1_1_ppc64le.whl", hash = "sha256:957b4774373cf6f709359e5c8c4a0af9f6d7875db657adb0feaf8d6cb3c3964c"},
{file = "yarl-1.9.4-cp37-cp37m-musllinux_1_1_s390x.whl", hash = "sha256:d7eeb6d22331e2fd42fce928a81c697c9ee2d51400bd1a28803965883e13cead"},
{file = "yarl-1.9.4-cp37-cp37m-musllinux_1_1_x86_64.whl", hash = "sha256:6a962e04b8f91f8c4e5917e518d17958e3bdee71fd1d8b88cdce74dd0ebbf434"},
{file = "yarl-1.9.4-cp37-cp37m-win32.whl", hash = "sha256:f3bc6af6e2b8f92eced34ef6a96ffb248e863af20ef4fde9448cc8c9b858b749"},
{file = "yarl-1.9.4-cp37-cp37m-win_amd64.whl", hash = "sha256:ad4d7a90a92e528aadf4965d685c17dacff3df282db1121136c382dc0b6014d2"},
{file = "yarl-1.9.4-cp38-cp38-macosx_10_9_universal2.whl", hash = "sha256:ec61d826d80fc293ed46c9dd26995921e3a82146feacd952ef0757236fc137be"},
{file = "yarl-1.9.4-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:8be9e837ea9113676e5754b43b940b50cce76d9ed7d2461df1af39a8ee674d9f"},
{file = "yarl-1.9.4-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:bef596fdaa8f26e3d66af846bbe77057237cb6e8efff8cd7cc8dff9a62278bbf"},
{file = "yarl-1.9.4-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:2d47552b6e52c3319fede1b60b3de120fe83bde9b7bddad11a69fb0af7db32f1"},
{file = "yarl-1.9.4-cp38-cp38-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:84fc30f71689d7fc9168b92788abc977dc8cefa806909565fc2951d02f6b7d57"},
{file = "yarl-1.9.4-cp38-cp38-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:4aa9741085f635934f3a2583e16fcf62ba835719a8b2b28fb2917bb0537c1dfa"},
{file = "yarl-1.9.4-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:206a55215e6d05dbc6c98ce598a59e6fbd0c493e2de4ea6cc2f4934d5a18d130"},
{file = "yarl-1.9.4-cp38-cp38-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:07574b007ee20e5c375a8fe4a0789fad26db905f9813be0f9fef5a68080de559"},
{file = "yarl-1.9.4-cp38-cp38-musllinux_1_1_aarch64.whl", hash = "sha256:5a2e2433eb9344a163aced6a5f6c9222c0786e5a9e9cac2c89f0b28433f56e23"},
{file = "yarl-1.9.4-cp38-cp38-musllinux_1_1_i686.whl", hash = "sha256:6ad6d10ed9b67a382b45f29ea028f92d25bc0bc1daf6c5b801b90b5aa70fb9ec"},
{file = "yarl-1.9.4-cp38-cp38-musllinux_1_1_ppc64le.whl", hash = "sha256:6fe79f998a4052d79e1c30eeb7d6c1c1056ad33300f682465e1b4e9b5a188b78"},
{file = "yarl-1.9.4-cp38-cp38-musllinux_1_1_s390x.whl", hash = "sha256:a825ec844298c791fd28ed14ed1bffc56a98d15b8c58a20e0e08c1f5f2bea1be"},
{file = "yarl-1.9.4-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:8619d6915b3b0b34420cf9b2bb6d81ef59d984cb0fde7544e9ece32b4b3043c3"},
{file = "yarl-1.9.4-cp38-cp38-win32.whl", hash = "sha256:686a0c2f85f83463272ddffd4deb5e591c98aac1897d65e92319f729c320eece"},
{file = "yarl-1.9.4-cp38-cp38-win_amd64.whl", hash = "sha256:a00862fb23195b6b8322f7d781b0dc1d82cb3bcac346d1e38689370cc1cc398b"},
{file = "yarl-1.9.4-cp39-cp39-macosx_10_9_universal2.whl", hash = "sha256:604f31d97fa493083ea21bd9b92c419012531c4e17ea6da0f65cacdcf5d0bd27"},
{file = "yarl-1.9.4-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:8a854227cf581330ffa2c4824d96e52ee621dd571078a252c25e3a3b3d94a1b1"},
{file = "yarl-1.9.4-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:ba6f52cbc7809cd8d74604cce9c14868306ae4aa0282016b641c661f981a6e91"},
{file = "yarl-1.9.4-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:a6327976c7c2f4ee6816eff196e25385ccc02cb81427952414a64811037bbc8b"},
{file = "yarl-1.9.4-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:8397a3817d7dcdd14bb266283cd1d6fc7264a48c186b986f32e86d86d35fbac5"},
{file = "yarl-1.9.4-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:e0381b4ce23ff92f8170080c97678040fc5b08da85e9e292292aba67fdac6c34"},
{file = "yarl-1.9.4-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:23d32a2594cb5d565d358a92e151315d1b2268bc10f4610d098f96b147370136"},
{file = "yarl-1.9.4-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:ddb2a5c08a4eaaba605340fdee8fc08e406c56617566d9643ad8bf6852778fc7"},
{file = "yarl-1.9.4-cp39-cp39-musllinux_1_1_aarch64.whl", hash = "sha256:26a1dc6285e03f3cc9e839a2da83bcbf31dcb0d004c72d0730e755b33466c30e"},
{file = "yarl-1.9.4-cp39-cp39-musllinux_1_1_i686.whl", hash = "sha256:18580f672e44ce1238b82f7fb87d727c4a131f3a9d33a5e0e82b793362bf18b4"},
{file = "yarl-1.9.4-cp39-cp39-musllinux_1_1_ppc64le.whl", hash = "sha256:29e0f83f37610f173eb7e7b5562dd71467993495e568e708d99e9d1944f561ec"},
{file = "yarl-1.9.4-cp39-cp39-musllinux_1_1_s390x.whl", hash = "sha256:1f23e4fe1e8794f74b6027d7cf19dc25f8b63af1483d91d595d4a07eca1fb26c"},
{file = "yarl-1.9.4-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:db8e58b9d79200c76956cefd14d5c90af54416ff5353c5bfd7cbe58818e26ef0"},
{file = "yarl-1.9.4-cp39-cp39-win32.whl", hash = "sha256:c7224cab95645c7ab53791022ae77a4509472613e839dab722a72abe5a684575"},
{file = "yarl-1.9.4-cp39-cp39-win_amd64.whl", hash = "sha256:824d6c50492add5da9374875ce72db7a0733b29c2394890aef23d533106e2b15"},
{file = "yarl-1.9.4-py3-none-any.whl", hash = "sha256:928cecb0ef9d5a7946eb6ff58417ad2fe9375762382f1bf5c55e61645f2c43ad"},
{file = "yarl-1.9.4.tar.gz", hash = "sha256:566db86717cf8080b99b58b083b773a908ae40f06681e87e589a976faf8246bf"},
]
[package.dependencies]

View File

@@ -87,6 +87,10 @@ impl AuthError {
pub fn too_many_connections() -> Self {
AuthErrorImpl::TooManyConnections.into()
}
pub fn is_auth_failed(&self) -> bool {
matches!(self.0.as_ref(), AuthErrorImpl::AuthFailed(_))
}
}
impl<E: Into<AuthErrorImpl>> From<E> for AuthError {

View File

@@ -9,7 +9,6 @@ use tokio_postgres::config::AuthKeys;
use crate::auth::credentials::check_peer_addr_is_in_list;
use crate::auth::validate_password_and_exchange;
use crate::console::errors::GetAuthInfoError;
use crate::console::provider::AuthInfo;
use crate::console::AuthSecret;
use crate::proxy::connect_compute::handle_try_wake;
use crate::proxy::retry::retry_after;
@@ -187,24 +186,52 @@ async fn auth_quirks(
};
info!("fetching user's authentication info");
// TODO(anna): this will slow down both "hacks" below; we probably need a cache.
let AuthInfo {
secret,
allowed_ips,
} = api.get_auth_info(extra, &info).await?;
let allowed_ips = api.get_allowed_ips(extra, &info).await?;
// check allowed list
if !check_peer_addr_is_in_list(&info.inner.peer_addr, &allowed_ips) {
return Err(auth::AuthError::ip_address_not_allowed());
}
let secret = secret.unwrap_or_else(|| {
let cached_secret = api.get_role_secret(extra, &info).await?;
let secret = cached_secret.clone().unwrap_or_else(|| {
// If we don't have an authentication secret, we mock one to
// prevent malicious probing (possible due to missing protocol steps).
// This mocked secret will never lead to successful authentication.
info!("authentication info not found, mocking it");
AuthSecret::Scram(scram::ServerSecret::mock(&info.inner.user, rand::random()))
});
match authenticate_with_secret(
secret,
info,
client,
unauthenticated_password,
allow_cleartext,
config,
latency_timer,
)
.await
{
Ok(keys) => Ok(keys),
Err(e) => {
if e.is_auth_failed() {
// The password could have been changed, so we invalidate the cache.
cached_secret.invalidate();
}
Err(e)
}
}
}
async fn authenticate_with_secret(
secret: AuthSecret,
info: ComputeUserInfo,
client: &mut stream::PqStream<Stream<impl AsyncRead + AsyncWrite + Unpin>>,
unauthenticated_password: Option<Vec<u8>>,
allow_cleartext: bool,
config: &'static AuthenticationConfig,
latency_timer: &mut LatencyTimer,
) -> auth::Result<ComputeCredentials<ComputeCredentialKeys>> {
if let Some(password) = unauthenticated_password {
let auth_outcome = validate_password_and_exchange(&password, secret)?;
let keys = match auth_outcome {

View File

@@ -6,6 +6,7 @@ use proxy::config::HttpConfig;
use proxy::console;
use proxy::console::provider::AllowedIpsCache;
use proxy::console::provider::NodeInfoCache;
use proxy::console::provider::RoleSecretCache;
use proxy::http;
use proxy::rate_limiter::EndpointRateLimiter;
use proxy::rate_limiter::RateBucketInfo;
@@ -86,7 +87,7 @@ struct ProxyCliArgs {
#[clap(long)]
metric_collection_interval: Option<String>,
/// cache for `wake_compute` api method (use `size=0` to disable)
#[clap(long, default_value = config::CacheOptions::DEFAULT_OPTIONS_NODE_INFO)]
#[clap(long, default_value = config::CacheOptions::CACHE_DEFAULT_OPTIONS)]
wake_compute_cache: String,
/// lock for `wake_compute` api method. example: "shards=32,permits=4,epoch=10m,timeout=1s". (use `permits=0` to disable).
#[clap(long, default_value = config::WakeComputeLockOptions::DEFAULT_OPTIONS_WAKE_COMPUTE_LOCK)]
@@ -127,8 +128,11 @@ struct ProxyCliArgs {
#[clap(flatten)]
aimd_config: proxy::rate_limiter::AimdConfig,
/// cache for `allowed_ips` (use `size=0` to disable)
#[clap(long, default_value = config::CacheOptions::DEFAULT_OPTIONS_NODE_INFO)]
#[clap(long, default_value = config::CacheOptions::CACHE_DEFAULT_OPTIONS)]
allowed_ips_cache: String,
/// cache for `role_secret` (use `size=0` to disable)
#[clap(long, default_value = config::CacheOptions::CACHE_DEFAULT_OPTIONS)]
role_secret_cache: String,
/// disable ip check for http requests. If it is too time consuming, it could be turned off.
#[clap(long, default_value_t = false, value_parser = clap::builder::BoolishValueParser::new(), action = clap::ArgAction::Set)]
disable_ip_check_for_http: bool,
@@ -266,9 +270,11 @@ fn build_config(args: &ProxyCliArgs) -> anyhow::Result<&'static ProxyConfig> {
AuthBackend::Console => {
let wake_compute_cache_config: CacheOptions = args.wake_compute_cache.parse()?;
let allowed_ips_cache_config: CacheOptions = args.allowed_ips_cache.parse()?;
let role_secret_cache_config: CacheOptions = args.role_secret_cache.parse()?;
info!("Using NodeInfoCache (wake_compute) with options={wake_compute_cache_config:?}");
info!("Using AllowedIpsCache (wake_compute) with options={allowed_ips_cache_config:?}");
info!("Using RoleSecretCache (wake_compute) with options={role_secret_cache_config:?}");
let caches = Box::leak(Box::new(console::caches::ApiCaches {
node_info: NodeInfoCache::new(
"node_info_cache",
@@ -282,6 +288,12 @@ fn build_config(args: &ProxyCliArgs) -> anyhow::Result<&'static ProxyConfig> {
allowed_ips_cache_config.ttl,
false,
),
role_secret: RoleSecretCache::new(
"role_secret_cache",
role_secret_cache_config.size,
role_secret_cache_config.ttl,
false,
),
}));
let config::WakeComputeLockOptions {

View File

@@ -310,10 +310,10 @@ pub struct CacheOptions {
impl CacheOptions {
/// Default options for [`crate::console::provider::NodeInfoCache`].
pub const DEFAULT_OPTIONS_NODE_INFO: &'static str = "size=4000,ttl=4m";
pub const CACHE_DEFAULT_OPTIONS: &'static str = "size=4000,ttl=4m";
/// Parse cache options passed via cmdline.
/// Example: [`Self::DEFAULT_OPTIONS_NODE_INFO`].
/// Example: [`Self::CACHE_DEFAULT_OPTIONS`].
fn parse(options: &str) -> anyhow::Result<Self> {
let mut size = None;
let mut ttl = None;

View File

@@ -10,6 +10,7 @@ use crate::{
};
use async_trait::async_trait;
use dashmap::DashMap;
use smol_str::SmolStr;
use std::{sync::Arc, time::Duration};
use tokio::{
sync::{OwnedSemaphorePermit, Semaphore},
@@ -216,6 +217,7 @@ impl ConsoleReqExtra {
}
/// Auth secret which is managed by the cloud.
#[derive(Clone)]
pub enum AuthSecret {
#[cfg(feature = "testing")]
/// Md5 hash of user's password.
@@ -250,18 +252,20 @@ pub struct NodeInfo {
pub type NodeInfoCache = TimedLru<Arc<str>, NodeInfo>;
pub type CachedNodeInfo = timed_lru::Cached<&'static NodeInfoCache>;
pub type AllowedIpsCache = TimedLru<Arc<str>, Arc<Vec<String>>>;
pub type AllowedIpsCache = TimedLru<SmolStr, Arc<Vec<String>>>;
pub type RoleSecretCache = TimedLru<(SmolStr, SmolStr), Option<AuthSecret>>;
pub type CachedRoleSecret = timed_lru::Cached<&'static RoleSecretCache>;
/// This will allocate per each call, but the http requests alone
/// already require a few allocations, so it should be fine.
#[async_trait]
pub trait Api {
/// Get the client's auth secret for authentication.
async fn get_auth_info(
async fn get_role_secret(
&self,
extra: &ConsoleReqExtra,
creds: &ComputeUserInfo,
) -> Result<AuthInfo, errors::GetAuthInfoError>;
) -> Result<CachedRoleSecret, errors::GetAuthInfoError>;
async fn get_allowed_ips(
&self,
@@ -282,7 +286,9 @@ pub struct ApiCaches {
/// Cache for the `wake_compute` API method.
pub node_info: NodeInfoCache,
/// Cache for the `get_allowed_ips`. TODO(anna): use notifications listener instead.
pub allowed_ips: TimedLru<Arc<str>, Arc<Vec<String>>>,
pub allowed_ips: AllowedIpsCache,
/// Cache for the `get_role_secret`. TODO(anna): use notifications listener instead.
pub role_secret: RoleSecretCache,
}
/// Various caches for [`console`](super).

View File

@@ -6,6 +6,7 @@ use super::{
errors::{ApiError, GetAuthInfoError, WakeComputeError},
AuthInfo, AuthSecret, CachedNodeInfo, ConsoleReqExtra, NodeInfo,
};
use crate::console::provider::CachedRoleSecret;
use crate::{auth::backend::ComputeUserInfo, compute, error::io_error, scram, url::ApiUrl};
use async_trait::async_trait;
use futures::TryFutureExt;
@@ -142,12 +143,14 @@ async fn get_execute_postgres_query(
#[async_trait]
impl super::Api for Api {
#[tracing::instrument(skip_all)]
async fn get_auth_info(
async fn get_role_secret(
&self,
_extra: &ConsoleReqExtra,
creds: &ComputeUserInfo,
) -> Result<AuthInfo, GetAuthInfoError> {
self.do_get_auth_info(creds).await
) -> Result<CachedRoleSecret, GetAuthInfoError> {
Ok(CachedRoleSecret::new_uncached(
self.do_get_auth_info(creds).await?.secret,
))
}
async fn get_allowed_ips(

View File

@@ -3,7 +3,8 @@
use super::{
super::messages::{ConsoleError, GetRoleSecret, WakeCompute},
errors::{ApiError, GetAuthInfoError, WakeComputeError},
ApiCaches, ApiLocks, AuthInfo, AuthSecret, CachedNodeInfo, ConsoleReqExtra, NodeInfo,
ApiCaches, ApiLocks, AuthInfo, AuthSecret, CachedNodeInfo, CachedRoleSecret, ConsoleReqExtra,
NodeInfo,
};
use crate::metrics::{ALLOWED_IPS_BY_CACHE_OUTCOME, ALLOWED_IPS_NUMBER};
use crate::{auth::backend::ComputeUserInfo, compute, http, scram};
@@ -159,12 +160,25 @@ impl Api {
#[async_trait]
impl super::Api for Api {
#[tracing::instrument(skip_all)]
async fn get_auth_info(
async fn get_role_secret(
&self,
extra: &ConsoleReqExtra,
creds: &ComputeUserInfo,
) -> Result<AuthInfo, GetAuthInfoError> {
self.do_get_auth_info(extra, creds).await
) -> Result<CachedRoleSecret, GetAuthInfoError> {
let ep = creds.endpoint.clone();
let user = creds.inner.user.clone();
if let Some(role_secret) = self.caches.role_secret.get(&(ep.clone(), user.clone())) {
return Ok(role_secret);
}
let auth_info = self.do_get_auth_info(extra, creds).await?;
let (_, secret) = self
.caches
.role_secret
.insert((ep.clone(), user), auth_info.secret.clone());
self.caches
.allowed_ips
.insert(ep, Arc::new(auth_info.allowed_ips));
Ok(secret)
}
async fn get_allowed_ips(
@@ -172,8 +186,7 @@ impl super::Api for Api {
extra: &ConsoleReqExtra,
creds: &ComputeUserInfo,
) -> Result<Arc<Vec<String>>, GetAuthInfoError> {
let key: &str = &creds.endpoint;
if let Some(allowed_ips) = self.caches.allowed_ips.get(key) {
if let Some(allowed_ips) = self.caches.allowed_ips.get(&creds.endpoint) {
ALLOWED_IPS_BY_CACHE_OUTCOME
.with_label_values(&["hit"])
.inc();
@@ -182,10 +195,14 @@ impl super::Api for Api {
ALLOWED_IPS_BY_CACHE_OUTCOME
.with_label_values(&["miss"])
.inc();
let allowed_ips = Arc::new(self.do_get_auth_info(extra, creds).await?.allowed_ips);
let auth_info = self.do_get_auth_info(extra, creds).await?;
let allowed_ips = Arc::new(auth_info.allowed_ips);
let ep = creds.endpoint.clone();
let user = creds.inner.user.clone();
self.caches
.allowed_ips
.insert(key.into(), allowed_ips.clone());
.role_secret
.insert((ep.clone(), user), auth_info.secret);
self.caches.allowed_ips.insert(ep, allowed_ips.clone());
Ok(allowed_ips)
}

View File

@@ -6,7 +6,7 @@ pub const SCRAM_KEY_LEN: usize = 32;
/// One of the keys derived from the [password](super::password::SaltedPassword).
/// We use the same structure for all keys, i.e.
/// `ClientKey`, `StoredKey`, and `ServerKey`.
#[derive(Default, PartialEq, Eq)]
#[derive(Clone, Default, PartialEq, Eq)]
#[repr(transparent)]
pub struct ScramKey {
bytes: [u8; SCRAM_KEY_LEN],

View File

@@ -5,6 +5,7 @@ use super::key::ScramKey;
/// Server secret is produced from [password](super::password::SaltedPassword)
/// and is used throughout the authentication process.
#[derive(Clone)]
pub struct ServerSecret {
/// Number of iterations for `PBKDF2` function.
pub iterations: u32,

View File

@@ -431,7 +431,6 @@ async fn connect_to_compute(
application_name: APP_NAME.to_string(),
options: console_options,
};
// TODO(anna): this is a bit hacky way, consider using console notification listener.
if !config.disable_ip_check_for_http {
let allowed_ips = backend.get_allowed_ips(&extra).await?;
if !check_peer_addr_is_in_list(&peer_addr, &allowed_ips) {

View File

@@ -1,25 +1,15 @@
# Collect /v1/debug_dump from all safekeeper nodes
1. Run ansible playbooks to collect .json dumps from all safekeepers and store them in `./result` directory.
3. Issue admin token (add/remove .stage from url for staging/prod and setting proper API key):
```
AUTH_TOKEN=$(curl https://console.stage.neon.tech/regions/console/api/v1/admin/issue_token -H "Accept: application/json" -H "Content-Type: application/json" -H "Authorization: Bearer $NEON_STAGING_KEY" -X POST -d '{"ttl_seconds": 43200, "scope": "safekeeperdata"}' 2>/dev/null | jq --raw-output '.jwt')
# check
echo $AUTH_TOKEN
```
2. Run ansible playbooks to collect .json dumps from all safekeepers and store them in `./result` directory.
```
# in aws repo, cd .github/ansible and run e.g. (ajusting profile and region in vars and limit):
AWS_DEFAULT_PROFILE=dev ansible-playbook -i inventory_aws_ec2.yaml -i staging.us-east-2.vars.yaml -e @ssm_config -l 'safekeeper:&us_east_2' -e "auth_token=${AUTH_TOKEN}" --check ~/neon/neon/scripts/sk_collect_dumps/remote.yaml
```
It will put the results to .results directory *near the playbook*.
2. Run `DB_CONNSTR=... ./upload.sh prod_feb30` to upload dumps to `prod_feb30` table in specified postgres database.
## How to use ansible (staging)
```
AWS_DEFAULT_PROFILE=dev ansible-playbook -i ../../.github/ansible/staging.us-east-2.hosts.yaml -e @../../.github/ansible/ssm_config remote.yaml
AWS_DEFAULT_PROFILE=dev ansible-playbook -i ../../.github/ansible/staging.eu-west-1.hosts.yaml -e @../../.github/ansible/ssm_config remote.yaml
```
## How to use ansible (prod)
```
AWS_DEFAULT_PROFILE=prod ansible-playbook -i ../../.github/ansible/prod.us-west-2.hosts.yaml -e @../../.github/ansible/ssm_config remote.yaml
AWS_DEFAULT_PROFILE=prod ansible-playbook -i ../../.github/ansible/prod.us-east-2.hosts.yaml -e @../../.github/ansible/ssm_config remote.yaml
AWS_DEFAULT_PROFILE=prod ansible-playbook -i ../../.github/ansible/prod.eu-central-1.hosts.yaml -e @../../.github/ansible/ssm_config remote.yaml
AWS_DEFAULT_PROFILE=prod ansible-playbook -i ../../.github/ansible/prod.ap-southeast-1.hosts.yaml -e @../../.github/ansible/ssm_config remote.yaml
```

View File

@@ -1,5 +1,5 @@
- name: Fetch state dumps from safekeepers
hosts: safekeepers
hosts: safekeeper
gather_facts: False
remote_user: "{{ remote_user }}"
@@ -8,6 +8,8 @@
get_url:
url: "http://{{ inventory_hostname }}:7676/v1/debug_dump?dump_all=true&dump_disk_content=false"
dest: "/tmp/{{ inventory_hostname }}.json"
headers:
Authorization: "Bearer {{ auth_token }}"
- name: Fetch file from remote hosts
fetch:

View File

@@ -31,22 +31,22 @@ SELECT
(data->>'tenant_id') AS tenant_id,
(data->>'timeline_id') AS timeline_id,
(data->'memory'->>'active')::bool AS active,
(data->'memory'->>'flush_lsn')::bigint AS flush_lsn,
(data->'memory'->'mem_state'->>'backup_lsn')::bigint AS backup_lsn,
(data->'memory'->'mem_state'->>'commit_lsn')::bigint AS commit_lsn,
(data->'memory'->'mem_state'->>'peer_horizon_lsn')::bigint AS peer_horizon_lsn,
(data->'memory'->'mem_state'->>'remote_consistent_lsn')::bigint AS remote_consistent_lsn,
(data->'memory'->>'write_lsn')::bigint AS write_lsn,
(data->'memory'->>'flush_lsn')::pg_lsn AS flush_lsn,
(data->'memory'->'mem_state'->>'backup_lsn')::pg_lsn AS backup_lsn,
(data->'memory'->'mem_state'->>'commit_lsn')::pg_lsn AS commit_lsn,
(data->'memory'->'mem_state'->>'peer_horizon_lsn')::pg_lsn AS peer_horizon_lsn,
(data->'memory'->'mem_state'->>'remote_consistent_lsn')::pg_lsn AS remote_consistent_lsn,
(data->'memory'->>'write_lsn')::pg_lsn AS write_lsn,
(data->'memory'->>'num_computes')::bigint AS num_computes,
(data->'memory'->>'epoch_start_lsn')::bigint AS epoch_start_lsn,
(data->'memory'->>'epoch_start_lsn')::pg_lsn AS epoch_start_lsn,
(data->'memory'->>'last_removed_segno')::bigint AS last_removed_segno,
(data->'memory'->>'is_cancelled')::bool AS is_cancelled,
(data->'control_file'->>'backup_lsn')::bigint AS disk_backup_lsn,
(data->'control_file'->>'commit_lsn')::bigint AS disk_commit_lsn,
(data->'control_file'->>'backup_lsn')::pg_lsn AS disk_backup_lsn,
(data->'control_file'->>'commit_lsn')::pg_lsn AS disk_commit_lsn,
(data->'control_file'->'acceptor_state'->>'term')::bigint AS disk_term,
(data->'control_file'->>'local_start_lsn')::bigint AS local_start_lsn,
(data->'control_file'->>'peer_horizon_lsn')::bigint AS disk_peer_horizon_lsn,
(data->'control_file'->>'timeline_start_lsn')::bigint AS timeline_start_lsn,
(data->'control_file'->>'remote_consistent_lsn')::bigint AS disk_remote_consistent_lsn
(data->'control_file'->>'local_start_lsn')::pg_lsn AS local_start_lsn,
(data->'control_file'->>'peer_horizon_lsn')::pg_lsn AS disk_peer_horizon_lsn,
(data->'control_file'->>'timeline_start_lsn')::pg_lsn AS timeline_start_lsn,
(data->'control_file'->>'remote_consistent_lsn')::pg_lsn AS disk_remote_consistent_lsn
FROM tmp_json
EOF

View File

@@ -510,13 +510,21 @@ class PageserverHttpClient(requests.Session):
assert res_json is None
def timeline_get_lsn_by_timestamp(
self, tenant_id: TenantId, timeline_id: TimelineId, timestamp, version: int
self,
tenant_id: TenantId,
timeline_id: TimelineId,
timestamp,
version: Optional[int] = None,
):
log.info(
f"Requesting lsn by timestamp {timestamp}, tenant {tenant_id}, timeline {timeline_id}"
)
if version is None:
version_str = ""
else:
version_str = f"&version={version}"
res = self.get(
f"http://localhost:{self.port}/v1/tenant/{tenant_id}/timeline/{timeline_id}/get_lsn_by_timestamp?timestamp={timestamp}&version={version}",
f"http://localhost:{self.port}/v1/tenant/{tenant_id}/timeline/{timeline_id}/get_lsn_by_timestamp?timestamp={timestamp}{version_str}",
)
self.verbose_error(res)
res_json = res.json()

View File

@@ -20,7 +20,7 @@ def test_local_corruption(neon_env_builder: NeonEnvBuilder):
env.pageserver.allowed_errors.extend(
[
".*layer loading failed:.*",
".*get_value_reconstruct_data for layer .*",
".*could not find data for key.*",
".*is not active. Current state: Broken.*",
".*will not become active. Current state: Broken.*",
@@ -83,7 +83,7 @@ def test_local_corruption(neon_env_builder: NeonEnvBuilder):
# (We don't check layer file contents on startup, when loading the timeline)
#
# This will change when we implement checksums for layers
with pytest.raises(Exception, match="layer loading failed:") as err:
with pytest.raises(Exception, match="get_value_reconstruct_data for layer ") as err:
pg2.start()
log.info(
f"As expected, compute startup failed for timeline {tenant2}/{timeline2} with corrupt layers: {err}"

View File

@@ -8,71 +8,6 @@ from fixtures.types import Lsn
from fixtures.utils import query_scalar
#
# Test pageserver get_lsn_by_timestamp API
#
def test_lsn_mapping_old(neon_env_builder: NeonEnvBuilder):
env = neon_env_builder.init_start()
new_timeline_id = env.neon_cli.create_branch("test_lsn_mapping")
endpoint_main = env.endpoints.create_start("test_lsn_mapping")
log.info("postgres is running on 'test_lsn_mapping' branch")
cur = endpoint_main.connect().cursor()
# Create table, and insert rows, each in a separate transaction
# Disable synchronous_commit to make this initialization go faster.
#
# Each row contains current insert LSN and the current timestamp, when
# the row was inserted.
cur.execute("SET synchronous_commit=off")
cur.execute("CREATE TABLE foo (x integer)")
tbl = []
for i in range(1000):
cur.execute("INSERT INTO foo VALUES(%s)", (i,))
# Get the timestamp at UTC
after_timestamp = query_scalar(cur, "SELECT clock_timestamp()").replace(tzinfo=None)
tbl.append([i, after_timestamp])
# Execute one more transaction with synchronous_commit enabled, to flush
# all the previous transactions
cur.execute("SET synchronous_commit=on")
cur.execute("INSERT INTO foo VALUES (-1)")
# Wait until WAL is received by pageserver
wait_for_last_flush_lsn(env, endpoint_main, env.initial_tenant, new_timeline_id)
with env.pageserver.http_client() as client:
# Check edge cases: timestamp in the future
probe_timestamp = tbl[-1][1] + timedelta(hours=1)
result = client.timeline_get_lsn_by_timestamp(
env.initial_tenant, new_timeline_id, f"{probe_timestamp.isoformat()}Z", 1
)
assert result == "future"
# timestamp too the far history
probe_timestamp = tbl[0][1] - timedelta(hours=10)
result = client.timeline_get_lsn_by_timestamp(
env.initial_tenant, new_timeline_id, f"{probe_timestamp.isoformat()}Z", 1
)
assert result == "past"
# Probe a bunch of timestamps in the valid range
for i in range(1, len(tbl), 100):
probe_timestamp = tbl[i][1]
lsn = client.timeline_get_lsn_by_timestamp(
env.initial_tenant, new_timeline_id, f"{probe_timestamp.isoformat()}Z", 1
)
# Call get_lsn_by_timestamp to get the LSN
# Launch a new read-only node at that LSN, and check that only the rows
# that were supposed to be committed at that point in time are visible.
endpoint_here = env.endpoints.create_start(
branch_name="test_lsn_mapping", endpoint_id="ep-lsn_mapping_read", lsn=lsn
)
assert endpoint_here.safe_psql("SELECT max(x) FROM foo")[0][0] == i
endpoint_here.stop_and_destroy()
#
# Test pageserver get_lsn_by_timestamp API
#
@@ -130,7 +65,7 @@ def test_lsn_mapping(neon_env_builder: NeonEnvBuilder):
# Timestamp is in the future
probe_timestamp = tbl[-1][1] + timedelta(hours=1)
result = client.timeline_get_lsn_by_timestamp(
tenant_id, timeline_id, f"{probe_timestamp.isoformat()}Z", 2
tenant_id, timeline_id, f"{probe_timestamp.isoformat()}Z"
)
assert result["kind"] == "future"
# make sure that we return a well advanced lsn here
@@ -139,7 +74,7 @@ def test_lsn_mapping(neon_env_builder: NeonEnvBuilder):
# Timestamp is in the unreachable past
probe_timestamp = tbl[0][1] - timedelta(hours=10)
result = client.timeline_get_lsn_by_timestamp(
tenant_id, timeline_id, f"{probe_timestamp.isoformat()}Z", 2
tenant_id, timeline_id, f"{probe_timestamp.isoformat()}Z"
)
assert result["kind"] == "past"
# make sure that we return the minimum lsn here at the start of the range
@@ -149,7 +84,7 @@ def test_lsn_mapping(neon_env_builder: NeonEnvBuilder):
for i in range(1, len(tbl), 100):
probe_timestamp = tbl[i][1]
result = client.timeline_get_lsn_by_timestamp(
tenant_id, timeline_id, f"{probe_timestamp.isoformat()}Z", 2
tenant_id, timeline_id, f"{probe_timestamp.isoformat()}Z"
)
assert result["kind"] not in ["past", "nodata"]
lsn = result["lsn"]