rust/neon
1
0
Fork 0
mirror of https://github.com/neondatabase/neon.git synced 2025-12-22 21:59:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fixes for enabling --timelines-onto-safekeepers in tests (#11854)

Browse Source 
Second PR with fixes extracted from #11712, relating to
`--timelines-onto-safekeepers`. Does the following:

* Moves safekeeper registration to `neon_local` instead of the test
fixtures
* Pass safekeeper JWT token if `--timelines-onto-safekeepers` is enabled
* Allow some warnings related to offline safekeepers (similarly to how
we allow them for offline pageservers)
* Enable generations on the compute's config if
`--timelines-onto-safekeepers` is enabled
* fix parallel `pull_timeline` race condition (the one that #11786 put
for later)

Fixes #11424
Part of #11670
This commit is contained in:
Arpad Müller
2025-05-08 17:13:11 +02:00
committed by GitHub
parent 659366060d
commit 622b3b2993
6 changed files with 131 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
safekeeper/src/http/routes.rs
View File

@@ -243,8 +243,7 @@ async fn timeline_pull_handler(mut request: Request<Body>) -> Result<Response<Bo
let resp =
pull_timeline::handle_request(data, conf.sk_auth_token.clone(), ca_certs, global_timelines)
.await
.map_err(ApiError::InternalServerError)?;
.await?;
json_response(StatusCode::OK, resp)
}

30
safekeeper/src/pull_timeline.rs
View File

@@ -7,6 +7,7 @@ use bytes::Bytes;
use camino::Utf8PathBuf;
use chrono::{DateTime, Utc};
use futures::{SinkExt, StreamExt, TryStreamExt};
use http_utils::error::ApiError;
use postgres_ffi::{PG_TLI, XLogFileName, XLogSegNo};
use reqwest::Certificate;
use safekeeper_api::Term;
@@ -30,7 +31,7 @@ use utils::pausable_failpoint;
use crate::control_file::CONTROL_FILE_NAME;
use crate::state::{EvictionState, TimelinePersistentState};
use crate::timeline::{Timeline, WalResidentTimeline};
use crate::timeline::{Timeline, TimelineError, WalResidentTimeline};
use crate::timelines_global_map::{create_temp_timeline_dir, validate_temp_timeline};
use crate::wal_storage::open_wal_file;
use crate::{GlobalTimelines, debug_dump, wal_backup};
@@ -395,7 +396,7 @@ pub async fn handle_request(
sk_auth_token: Option<SecretString>,
ssl_ca_certs: Vec<Certificate>,
global_timelines: Arc<GlobalTimelines>,
) -> Result<PullTimelineResponse> {
) -> Result<PullTimelineResponse, ApiError> {
let existing_tli = global_timelines.get(TenantTimelineId::new(
request.tenant_id,
request.timeline_id,
@@ -411,7 +412,9 @@ pub async fn handle_request(
for ssl_ca_cert in ssl_ca_certs {
http_client = http_client.add_root_certificate(ssl_ca_cert);
}
let http_client = http_client.build()?;
let http_client = http_client
.build()
.map_err(|e| ApiError::InternalServerError(e.into()))?;
let http_hosts = request.http_hosts.clone();
@@ -443,10 +446,10 @@ pub async fn handle_request(
// offline and C comes online. Then we want a pull on C with A and B as hosts to work.
let min_required_successful = (http_hosts.len() - 1).max(1);
if statuses.len() < min_required_successful {
bail!(
return Err(ApiError::InternalServerError(anyhow::anyhow!(
"only got {} successful status responses. required: {min_required_successful}",
statuses.len()
)
)));
}
// Find the most advanced safekeeper
@@ -465,7 +468,7 @@ pub async fn handle_request(
assert!(status.tenant_id == request.tenant_id);
assert!(status.timeline_id == request.timeline_id);
pull_timeline(
match pull_timeline(
status,
safekeeper_host,
sk_auth_token,
@@ -473,6 +476,21 @@ pub async fn handle_request(
global_timelines,
)
.await
{
Ok(resp) => Ok(resp),
Err(e) => {
match e.downcast_ref::<TimelineError>() {
Some(TimelineError::AlreadyExists(_)) => Ok(PullTimelineResponse {
safekeeper_host: None,
}),
Some(TimelineError::CreationInProgress(_)) => {
// We don't return success here because creation might still fail.
Err(ApiError::Conflict("Creation in progress".to_owned()))
}
_ => Err(ApiError::InternalServerError(e)),
}
}
}
}
async fn pull_timeline(