Object storage proxy (#11357)

Service targeted for storing and retrieving LFC prewarm data.
Can be used for proxying S3 access for Postgres extensions like
pg_mooncake as well.

Requests must include a Bearer JWT token.
Token is validated using a pemfile (should be passed in infra/).

Note: app is not tolerant to extra trailing slashes, see app.rs
`delete_prefix` test for comments.

Resolves: https://github.com/neondatabase/cloud/issues/26342
Unrelated changes: gate a `rename_noreplace` feature and disable it in
`remote_storage` so as `object_storage` can be built with musl
This commit is contained in:
Mikhail Kot
2025-04-08 15:54:53 +01:00
committed by GitHub
parent a7142f3bc6
commit 6138d61592
23 changed files with 1424 additions and 38 deletions

View File

@@ -20,8 +20,10 @@ use compute_api::spec::ComputeMode;
use control_plane::endpoint::ComputeControlPlane;
use control_plane::local_env::{
InitForceMode, LocalEnv, NeonBroker, NeonLocalInitConf, NeonLocalInitPageserverConf,
SafekeeperConf,
ObjectStorageConf, SafekeeperConf,
};
use control_plane::object_storage::OBJECT_STORAGE_DEFAULT_PORT;
use control_plane::object_storage::ObjectStorage;
use control_plane::pageserver::PageServerNode;
use control_plane::safekeeper::SafekeeperNode;
use control_plane::storage_controller::{
@@ -91,6 +93,8 @@ enum NeonLocalCmd {
#[command(subcommand)]
Safekeeper(SafekeeperCmd),
#[command(subcommand)]
ObjectStorage(ObjectStorageCmd),
#[command(subcommand)]
Endpoint(EndpointCmd),
#[command(subcommand)]
Mappings(MappingsCmd),
@@ -454,6 +458,32 @@ enum SafekeeperCmd {
Restart(SafekeeperRestartCmdArgs),
}
#[derive(clap::Subcommand)]
#[clap(about = "Manage object storage")]
enum ObjectStorageCmd {
Start(ObjectStorageStartCmd),
Stop(ObjectStorageStopCmd),
}
#[derive(clap::Args)]
#[clap(about = "Start object storage")]
struct ObjectStorageStartCmd {
#[clap(short = 't', long, help = "timeout until we fail the command")]
#[arg(default_value = "10s")]
start_timeout: humantime::Duration,
}
#[derive(clap::Args)]
#[clap(about = "Stop object storage")]
struct ObjectStorageStopCmd {
#[arg(value_enum, default_value = "fast")]
#[clap(
short = 'm',
help = "If 'immediate', don't flush repository data at shutdown"
)]
stop_mode: StopMode,
}
#[derive(clap::Args)]
#[clap(about = "Start local safekeeper")]
struct SafekeeperStartCmdArgs {
@@ -759,6 +789,7 @@ fn main() -> Result<()> {
}
NeonLocalCmd::StorageBroker(subcmd) => rt.block_on(handle_storage_broker(&subcmd, env)),
NeonLocalCmd::Safekeeper(subcmd) => rt.block_on(handle_safekeeper(&subcmd, env)),
NeonLocalCmd::ObjectStorage(subcmd) => rt.block_on(handle_object_storage(&subcmd, env)),
NeonLocalCmd::Endpoint(subcmd) => rt.block_on(handle_endpoint(&subcmd, env)),
NeonLocalCmd::Mappings(subcmd) => handle_mappings(&subcmd, env),
};
@@ -975,6 +1006,9 @@ fn handle_init(args: &InitCmdArgs) -> anyhow::Result<LocalEnv> {
}
})
.collect(),
object_storage: ObjectStorageConf {
port: OBJECT_STORAGE_DEFAULT_PORT,
},
pg_distrib_dir: None,
neon_distrib_dir: None,
default_tenant_id: TenantId::from_array(std::array::from_fn(|_| 0)),
@@ -1683,6 +1717,41 @@ async fn handle_safekeeper(subcmd: &SafekeeperCmd, env: &local_env::LocalEnv) ->
Ok(())
}
async fn handle_object_storage(subcmd: &ObjectStorageCmd, env: &local_env::LocalEnv) -> Result<()> {
use ObjectStorageCmd::*;
let storage = ObjectStorage::from_env(env);
// In tests like test_forward_compatibility or test_graceful_cluster_restart
// old neon binaries (without object_storage) are present
if !storage.bin.exists() {
eprintln!(
"{} binary not found. Ignore if this is a compatibility test",
storage.bin
);
return Ok(());
}
match subcmd {
Start(ObjectStorageStartCmd { start_timeout }) => {
if let Err(e) = storage.start(start_timeout).await {
eprintln!("object_storage start failed: {e}");
exit(1);
}
}
Stop(ObjectStorageStopCmd { stop_mode }) => {
let immediate = match stop_mode {
StopMode::Fast => false,
StopMode::Immediate => true,
};
if let Err(e) = storage.stop(immediate) {
eprintln!("proxy stop failed: {e}");
exit(1);
}
}
};
Ok(())
}
async fn handle_storage_broker(subcmd: &StorageBrokerCmd, env: &local_env::LocalEnv) -> Result<()> {
match subcmd {
StorageBrokerCmd::Start(args) => {
@@ -1777,6 +1846,13 @@ async fn handle_start_all_impl(
.map_err(|e| e.context(format!("start safekeeper {}", safekeeper.id)))
});
}
js.spawn(async move {
ObjectStorage::from_env(env)
.start(&retry_timeout)
.await
.map_err(|e| e.context("start object_storage"))
});
})();
let mut errors = Vec::new();
@@ -1874,6 +1950,11 @@ async fn try_stop_all(env: &local_env::LocalEnv, immediate: bool) {
}
}
let storage = ObjectStorage::from_env(env);
if let Err(e) = storage.stop(immediate) {
eprintln!("object_storage stop failed: {:#}", e);
}
for ps_conf in &env.pageservers {
let pageserver = PageServerNode::from_env(env, ps_conf);
if let Err(e) = pageserver.stop(immediate) {