rust/neon
1
0
Fork 0
mirror of https://github.com/neondatabase/neon.git synced 2026-05-27 18:10:37 +00:00
Code Issues Packages Projects Releases Wiki Activity

proxy: Demote all cplane error replies to info log level (#9880)

Browse Source 
## Problem

The vast majority of the error/warn logs from cplane are about time or
data transfer quotas exceeded or endpoint-not-found errors and not
operational errors in proxy or cplane.

## Summary of changes

* Demote cplane error replies to info level.
* Raise other errors from warn back to error.
This commit is contained in:
Folke Behrens
2024-11-25 18:53:26 +01:00
committed by GitHub
parent 7a2f0ed8d4
commit 87e4dd23a1
1 changed files with 16 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
proxy/src/proxy/wake_compute.rs
View File

@@ -1,9 +1,9 @@
use tracing::{error, info, warn};
use tracing::{error, info};
use super::connect_compute::ComputeConnectBackend;
use crate::config::RetryConfig;
use crate::context::RequestContext;
use crate::control_plane::errors::WakeComputeError;
use crate::control_plane::errors::{ControlPlaneError, WakeComputeError};
use crate::control_plane::CachedNodeInfo;
use crate::error::ReportableError;
use crate::metrics::{
@@ -11,6 +11,18 @@ use crate::metrics::{
};
use crate::proxy::retry::{retry_after, should_retry};
// Use macro to retain original callsite.
macro_rules! log_wake_compute_error {
(error = ?$error:expr, $num_retries:expr, retriable = $retriable:literal) => {
match $error {
WakeComputeError::ControlPlane(ControlPlaneError::Message(_)) => {
info!(error = ?$error, num_retries = $num_retries, retriable = $retriable, "couldn't wake compute node")
}
_ => error!(error = ?$error, num_retries = $num_retries, retriable = $retriable, "couldn't wake compute node"),
}
};
}
pub(crate) async fn wake_compute<B: ComputeConnectBackend>(
num_retries: &mut u32,
ctx: &RequestContext,
@@ -20,7 +32,7 @@ pub(crate) async fn wake_compute<B: ComputeConnectBackend>(
loop {
match api.wake_compute(ctx).await {
Err(e) if !should_retry(&e, *num_retries, config) => {
error!(error = ?e, num_retries, retriable = false, "couldn't wake compute node");
log_wake_compute_error!(error = ?e, num_retries, retriable = false);
report_error(&e, false);
Metrics::get().proxy.retries_metric.observe(
RetriesMetricGroup {
@@ -32,7 +44,7 @@ pub(crate) async fn wake_compute<B: ComputeConnectBackend>(
return Err(e);
}
Err(e) => {
warn!(error = ?e, num_retries, retriable = true, "couldn't wake compute node");
log_wake_compute_error!(error = ?e, num_retries, retriable = true);
report_error(&e, true);
}
Ok(n) => {