You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
feat(envd): add mem_cache metric to track page cache usage (#2269)
* feat(envd): add mem_cache metric to track page cache usage across full pipeline
Expose the Linux page cache (Cached from /proc/meminfo) as a new mem_cache
metric. Previously only mem_used was reported, which excludes page cache and
can underreport actual memory pressure inside Firecracker VMs.
The metric flows through: envd -> orchestrator OTel gauges -> ClickHouse ->
public and edge APIs. Version-gated to envd >= 0.5.9 for backward compat.
* fix(api): make memCache a required field in SandboxMetric schemas
All other metric fields (memUsed, memTotal, diskUsed, etc.) are required in
the OpenAPI specs. Leaving memCache optional caused oapi-codegen to generate
it as a pointer (*int64) with omitempty, which is inconsistent with the rest
of the struct and would make it the only metric that could be absent from API
responses.
* chore: auto-commit generated changes
* fix(telemetry): add description and unit for ram cache gauge
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
0 commit comments