mirror of
https://github.com/rcourtman/Pulse.git
synced 2026-10-23 20:25:56 +00:00
Keep used-memory history within its retention window
Periodic guest and node cleanup omitted raw memory-used samples from both expiry and live-entry checks. Trim them with the existing retention and backing-array policy without deleting a still-live byte series. Cover every supported History series, byte-array reclamation, concurrent cleanup and poll-to-chart/metric-window resumption. No SQL retention, source-selection or wire changes. Change-source: pulse-maintainer
This commit is contained in:
@@ -17,6 +17,20 @@
|
||||
|
||||
## Purpose
|
||||
|
||||
### Used-memory history retention
|
||||
|
||||
Periodic in-memory History cleanup applies the existing retention window and
|
||||
backing-array release policy to raw `memoryused` bytes as well as `memory`
|
||||
percentages, for guests and nodes. A live byte series alone keeps its resource
|
||||
entry; an expired byte series cannot remain pinned by ongoing CPU or other
|
||||
observations. Missing current memory does not refresh its last-known samples.
|
||||
The existing poll recorder, chart readers and metric-window provider retain
|
||||
in-window byte values and their original timestamps through cleanup, then
|
||||
accept fresh memory on resumption. SQL retention, sampling limits, memory
|
||||
source selection and wire formats are unchanged. The History cleanup tests
|
||||
cover every supported series, byte-array reclamation and connected readers;
|
||||
these controls do not attribute a reporter's native process-memory growth.
|
||||
|
||||
### Podman zero-percentage fallback — issue #2447
|
||||
|
||||
One-shot compatibility stats with `cpu:0` do not suppress advancing cumulative
|
||||
|
||||
Reference in New Issue
Block a user