[proxy] add metrics for caches (#12752)

Exposes metrics for caches. LKB-2594

This exposes a high level namespace, `cache`, that all cache metrics can
be added to - this makes it easier to make library panels for the caches
as I understand it.

To calculate the current cache fill ratio, you could use the following
query:

```
(
    cache_inserted_total{cache="node_info"}
  - sum (cache_evicted_total{cache="node_info"}) without (cause)
)
  / cache_capacity{cache="node_info"}
```

To calculate the cache hit ratio, you could use the following query:

```
  cache_request_total{cache="node_info", outcome="hit"}
/ sum (cache_request_total{cache="node_info"}) without (outcome)
```
This commit is contained in:
Conrad Ludgate
2025-07-28 11:41:49 +01:00
committed by GitHub
parent a6e0baf31a
commit effd6bf829
6 changed files with 177 additions and 29 deletions

View File

@@ -1,5 +1,6 @@
use std::borrow::Cow;
use std::collections::HashMap;
use std::convert::Infallible;
use std::sync::Arc;
use bytes::Bytes;
@@ -54,11 +55,12 @@ use super::http_util::{
};
use super::json::JsonConversionError;
use crate::auth::backend::ComputeCredentialKeys;
use crate::cache::common::{count_cache_insert, count_cache_outcome, eviction_listener};
use crate::config::ProxyConfig;
use crate::context::RequestContext;
use crate::error::{ErrorKind, ReportableError, UserFacingError};
use crate::http::read_body_with_limit;
use crate::metrics::Metrics;
use crate::metrics::{CacheKind, Metrics};
use crate::serverless::sql_over_http::HEADER_VALUE_TRUE;
use crate::types::EndpointCacheKey;
use crate::util::deserialize_json_string;
@@ -138,15 +140,31 @@ pub struct ApiConfig {
}
// The DbSchemaCache is a cache of the ApiConfig and DbSchemaOwned for each endpoint
pub(crate) struct DbSchemaCache(pub Cache<EndpointCacheKey, Arc<(ApiConfig, DbSchemaOwned)>>);
pub(crate) struct DbSchemaCache(Cache<EndpointCacheKey, Arc<(ApiConfig, DbSchemaOwned)>>);
impl DbSchemaCache {
pub fn new(config: crate::config::CacheOptions) -> Self {
let builder = Cache::builder().name("db_schema_cache");
let builder = Cache::builder().name("schema");
let builder = config.moka(builder);
let metrics = &Metrics::get().cache;
if let Some(size) = config.size {
metrics.capacity.set(CacheKind::Schema, size as i64);
}
let builder =
builder.eviction_listener(|_k, _v, cause| eviction_listener(CacheKind::Schema, cause));
Self(builder.build())
}
pub async fn maintain(&self) -> Result<Infallible, anyhow::Error> {
let mut ticker = tokio::time::interval(std::time::Duration::from_secs(60));
loop {
ticker.tick().await;
self.0.run_pending_tasks();
}
}
pub async fn get_cached_or_remote(
&self,
endpoint_id: &EndpointCacheKey,
@@ -156,7 +174,8 @@ impl DbSchemaCache {
ctx: &RequestContext,
config: &'static ProxyConfig,
) -> Result<Arc<(ApiConfig, DbSchemaOwned)>, RestError> {
match self.0.get(endpoint_id) {
let cache_result = count_cache_outcome(CacheKind::Schema, self.0.get(endpoint_id));
match cache_result {
Some(v) => Ok(v),
None => {
info!("db_schema cache miss for endpoint: {:?}", endpoint_id);
@@ -180,6 +199,7 @@ impl DbSchemaCache {
db_extra_search_path: None,
};
let value = Arc::new((api_config, schema_owned));
count_cache_insert(CacheKind::Schema);
self.0.insert(endpoint_id.clone(), value);
return Err(e);
}
@@ -188,6 +208,7 @@ impl DbSchemaCache {
}
};
let value = Arc::new((api_config, schema_owned));
count_cache_insert(CacheKind::Schema);
self.0.insert(endpoint_id.clone(), value.clone());
Ok(value)
}