j1wonpark opened a new issue, #4313: URL: https://github.com/apache/amoro/issues/4313
### Parent issue #4191 (AIP-5: Dynamic Resource Allocation for Optimizer) ### Description Final phase of [AIP-5](https://cwiki.apache.org/confluence/display/AMORO/AIP-5%3A+Dynamic+Resource+Allocation+for+Optimizer): expose the scaling behavior of dynamic allocation groups as per-group metrics (all tagged `group`). **Gauges** - `optimizer_group_idle_optimizers` — instances with zero in-flight tasks - `optimizer_group_config_invalid` — `1` while an opted-in group's dynamic allocation config is invalid and the startup fail-safe keeps it disabled - `optimizer_group_pending_removal_optimizers` — instances in graceful drain - `optimizer_group_effective_threads` — registered threads plus threads of optimizers pending registration - `optimizer_group_backlog_duration_ms` — duration since demand first exceeded capacity **Counters** - `optimizer_group_scale_up_total` / `optimizer_group_scale_down_total` — cumulative scaling actions, one per scale-out round / drain start, so thrashing-style alerts become expressible with `increase(...[1h])` The first two gauges register with the existing `OptimizerGroupMetrics` for every group; the other five live with the scale keeper's watch on the group, since their sources (drain set, boot-window accounting, backlog timer) exist only while dynamic allocation is enabled. Also documents the new metrics in `docs/user-guides/metrics.md`. Depends on #4295. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
