FEATURE: all VictoriaMetrics Enterprise components: run with the -eula command-line flag to accept ESA. License keys are no longer verified. The -license, -licenseFile, -licenseFile.reloadInterval and -license.forceOffline flags are deprecated and are treated as -eula, so existing setups keep working. See Running VictoriaMetrics Enterprise.
FEATURE: all VictoriaMetrics components: add new metric process_cgroup_cpu_throttled_seconds_total which exposes CPU throttling. See #10430.
FEATURE: vmagent: add a new flag -remoteWrite.queuePath, which allows configuring persistent-queue path for each -remoteWrite.url independently. See #11662.
FEATURE: vmagent: add vmagent_kafka_consumer_rebalance_events_total and vmagent_kafka_consumer_assigned_partitions metrics for monitoring partition rebalancing and distribution. See partition distribution and #11370.
FEATURE: vmalert: add -notifier.sanitizeLabelNames command-line flag. Set it to false to send alert label names to notifiers without replacing characters outside [a-zA-Z0-9_:] with underscores. This is useful for Alertmanager v0.27.0 or newer, which supports UTF-8 label names when not running in classic mode. See #11629.
FEATURE: vmsingle, vmstorage and vmselect in VictoriaMetrics cluster: expose the working-set cache mode via vm_cache_mode for storage/tsid, storage/metricIDs, storage/metricName, indexdb/tagFiltersLoops, and promql/rollupResult cache. Values are 0 for split, 1 for switching, and 2 for whole. This metric helps users understand the internal memory and cache mechanisms. See #11860.
BUGFIX: all VictoriaMetrics components: OS solaris fixes compilation error. See #11850. Thanks to @szaydel for contribution.
BUGFIX: all VictoriaMetrics components: correctly detect CPU and memorycgroup limits configured via systemd slices in cgroup v1. Previously, only the global host cgroup limit was used, which could cause OOM errors and CPU throttling. See #11827.
BUGFIX: alerts: stop vmbackupmanager alerting rulesNoLatestBackupWithinLastDay, NoHourlyBackupWithinLastDay, NoDailyBackupWithinLast3Days, NoWeeklyBackupWithinLast14Days and NoMonthlyBackupWithinLast62Days from firing when scheduled backups are disabled via -disableScheduledBackups or the corresponding backup type is disabled via -disableHourly, -disableDaily, -disableWeekly or -disableMonthly. See #11543.
BUGFIX: stream aggregation: properly delay flushes after a sudden ingestion lag spike when aggregation windows are enabled. Previously, a data gap or a restart could reduce the tolerance for delayed samples to the aggregation interval, resulting in more dropped samples than necessary. See #11866.
BUGFIX: stream aggregation: prevent skipping aggregation and deduplication flushes when aggregation windows are enabled and incoming samples are delayed by one interval or more, e.g. when the ingestion pipeline is overloaded. Previously, the flush delay could grow with the sample lag, causing skipped flushes and gaps in the results. Now the flush delay is capped at 90% of the aggregation or deduplication interval. See #11811.
BUGFIX: stream aggregation: prevent spikes in rate_sum and rate_avg outputs with enable_windows: true when out-of-order samples arrive around the aggregation window boundary. Previously, an out-of-order sample could be treated as a counter reset, causing large spike in the result. See #11810.
BUGFIX: vmalert: send staleness markers for recording rule series with their original labels when a label value contains , or =, and avoid reporting distinct label sets as duplicates when these symbols occur in label values. Previously, staleness markers could be sent with broken labels and distinct label sets could collide in the series key. See #11851. Thanks to @sarankumar1325 for contribution.
BUGFIX: vmsingle and vmselect in VictoriaMetrics cluster: properly calculate running_avg and range_avg for series with NaN points in the middle. Previously, every NaN point was counted in the divisor, so each value after the gap moved the result by less than it should. See #11823. Thanks to @spicysharkco for contribution.
BUGFIX: vmsingle and vmselect in VictoriaMetrics cluster: do not increase vm_http_request_errors_total when a query is cancelled. This prevents false RequestErrorsToAPI alerts from firing. See #11828.
BUGFIX: vmsingle and vmselect in VictoriaMetrics cluster: correctly cancel query request if -search.skipSlowReplicas flag is set. See #11771 for details.
BUGFIX: vmui: fixes predefined dashboard rows not expanding when clicking the row title. See #11829. Thanks to @nikshrma for contribution.
BUGFIX: vmui: support label names with special characters in series aliases. See #11822. Thanks to @kema-dev for contribution.
Configuration
📅Schedule: (UTC)
Branch creation
At any time (no schedule defined)
Automerge
At any time (no schedule defined)
🚦Automerge: Disabled by config. Please merge this manually once you are satisfied.
♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.
🔕Ignore: Close this PR and you won't be reminded about this update again.
If you want to rebase/retry this PR, check this box
This PR contains the following updates:
| Package | Update | Change |
|---|---|---|
| [victoriametrics/vmagent](https://github.com/VictoriaMetrics/VictoriaMetrics) | minor | `v1.155.0` → `v1.156.0` |
---
### Release Notes
<details>
<summary>VictoriaMetrics/VictoriaMetrics (victoriametrics/vmagent)</summary>
### [`v1.156.0`](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.156.0)
[Compare Source](https://github.com/VictoriaMetrics/VictoriaMetrics/compare/v1.155.0...v1.156.0)
##### [v1.156.0](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.156.0)
Released at 2026-11-09
- FEATURE: all VictoriaMetrics [Enterprise](https://docs.victoriametrics.com/victoriametrics/enterprise/) components: run with the `-eula` command-line flag to accept [ESA](https://victoriametrics.com/legal/esa/). License keys are no longer verified. The `-license`, `-licenseFile`, `-licenseFile.reloadInterval` and `-license.forceOffline` flags are deprecated and are treated as `-eula`, so existing setups keep working. See [Running VictoriaMetrics Enterprise](https://docs.victoriametrics.com/victoriametrics/enterprise/#running-victoriametrics-enterprise).
- FEATURE: all VictoriaMetrics components: add new metric `process_cgroup_cpu_throttled_seconds_total` which exposes CPU throttling. See [#​10430](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10430).
- FEATURE: [vmagent](https://docs.victoriametrics.com/victoriametrics/vmagent/) and [Single-node VictoriaMetrics](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/): support `authentication_method: WorkloadIdentity` in [azure_sd_configs](https://docs.victoriametrics.com/victoriametrics/sd_configs/#azure_sd_configs). It allows discovering Azure VMs from Kubernetes clusters with [workload identity federation](https://learn.microsoft.com/en-us/entra/workload-id/workload-identity-federation) without a static client secret. See [#​11558](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11558). Thanks to [@​banlor](https://github.com/banlor) for contribution.
- FEATURE: [vmagent](https://docs.victoriametrics.com/victoriametrics/vmagent/): add a new flag `-remoteWrite.queuePath`, which allows configuring [persistent-queue](https://docs.victoriametrics.com/victoriametrics/vmagent/#on-disk-persistence) path for each `-remoteWrite.url` independently. See [#​11662](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11662).
- FEATURE: [vmagent](https://docs.victoriametrics.com/victoriametrics/vmagent/): add `vmagent_kafka_consumer_rebalance_events_total` and `vmagent_kafka_consumer_assigned_partitions` metrics for monitoring partition rebalancing and distribution. See [partition distribution](https://docs.victoriametrics.com/victoriametrics/integrations/kafka/#partition-distribution) and [#​11370](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11370).
- FEATURE: [vmalert](https://docs.victoriametrics.com/victoriametrics/vmalert/): add `-notifier.sanitizeLabelNames` command-line flag. Set it to `false` to send alert label names to notifiers without replacing characters outside `[a-zA-Z0-9_:]` with underscores. This is useful for Alertmanager v0.27.0 or newer, which [supports UTF-8 label names](https://prometheus.io/docs/alerting/latest/configuration/#utf-8-strict-mode) when not running in classic mode. See [#​11629](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11629).
- FEATURE: [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/), `vmstorage` and `vmselect` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): expose the working-set cache mode via `vm_cache_mode` for `storage/tsid`, `storage/metricIDs`, `storage/metricName`, `indexdb/tagFiltersLoops`, and `promql/rollupResult` cache. Values are `0` for split, `1` for switching, and `2` for whole. This metric helps users understand the internal memory and cache mechanisms. See [#​11860](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11860).
- BUGFIX: all VictoriaMetrics components: OS `solaris` fixes compilation error. See [#​11850](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/11850). Thanks to [@​szaydel](https://github.com/szaydel) for contribution.
- BUGFIX: all VictoriaMetrics components: correctly detect `CPU` and `memory` `cgroup` limits configured via [systemd slices](https://www.freedesktop.org/software/systemd/man/latest/systemd.slice.html) in `cgroup v1`. Previously, only the global host cgroup limit was used, which could cause OOM errors and CPU throttling. See [#​11827](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11827).
- BUGFIX: [alerts](https://github.com/VictoriaMetrics/VictoriaMetrics/blob/master/deployment/docker/rules): stop [vmbackupmanager alerting rules](https://github.com/VictoriaMetrics/VictoriaMetrics/blob/master/deployment/docker/rules/alerts-vmbackupmanager.yml) `NoLatestBackupWithinLastDay`, `NoHourlyBackupWithinLastDay`, `NoDailyBackupWithinLast3Days`, `NoWeeklyBackupWithinLast14Days` and `NoMonthlyBackupWithinLast62Days` from firing when scheduled backups are disabled via `-disableScheduledBackups` or the corresponding backup type is disabled via `-disableHourly`, `-disableDaily`, `-disableWeekly` or `-disableMonthly`. See [#​11543](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11543).
- BUGFIX: [stream aggregation](https://docs.victoriametrics.com/victoriametrics/stream-aggregation/): properly delay flushes after a sudden ingestion lag spike when [aggregation windows](https://docs.victoriametrics.com/victoriametrics/stream-aggregation/#aggregation-windows) are enabled. Previously, a data gap or a restart could reduce the tolerance for delayed samples to the aggregation interval, resulting in more dropped samples than necessary. See [#​11866](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/11866).
- BUGFIX: [stream aggregation](https://docs.victoriametrics.com/victoriametrics/stream-aggregation/): prevent skipping aggregation and deduplication flushes when [aggregation windows](https://docs.victoriametrics.com/victoriametrics/stream-aggregation/#aggregation-windows) are enabled and incoming samples are delayed by one interval or more, e.g. when the ingestion pipeline is overloaded. Previously, the flush delay could grow with the sample lag, causing skipped flushes and gaps in the results. Now the flush delay is capped at 90% of the aggregation or deduplication interval. See [#​11811](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11811).
- BUGFIX: [stream aggregation](https://docs.victoriametrics.com/victoriametrics/stream-aggregation/): prevent spikes in `rate_sum` and `rate_avg` outputs with `enable_windows: true` when out-of-order samples arrive around the aggregation window boundary. Previously, an out-of-order sample could be treated as a counter reset, causing large spike in the result. See [#​11810](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11810).
- BUGFIX: [vmalert](https://docs.victoriametrics.com/victoriametrics/vmalert/): send staleness markers for recording rule series with their original labels when a label value contains `,` or `=`, and avoid reporting distinct label sets as duplicates when these symbols occur in label values. Previously, staleness markers could be sent with broken labels and distinct label sets could collide in the series key. See [#​11851](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/11851). Thanks to [@​sarankumar1325](https://github.com/sarankumar1325) for contribution.
- BUGFIX: [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmselect` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): properly calculate [running_avg](https://docs.victoriametrics.com/victoriametrics/metricsql/#running_avg) and [range_avg](https://docs.victoriametrics.com/victoriametrics/metricsql/#range_avg) for series with `NaN` points in the middle. Previously, every `NaN` point was counted in the divisor, so each value after the gap moved the result by less than it should. See [#​11823](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/11823). Thanks to [@​spicysharkco](https://github.com/spicysharkco) for contribution.
- BUGFIX: [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmselect` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): do not increase `vm_http_request_errors_total` when a query is cancelled. This prevents false `RequestErrorsToAPI` alerts from firing. See [#​11828](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11828).
- BUGFIX: [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmselect` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): properly parse `metricsql` expression numbers with leading zeroes, such as `09` or `08`. See [#​11621](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11621). Thanks to [@​andriibeee](https://github.com/andriibeee) for contribution.
- BUGFIX: [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmselect` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): properly check `-search.maxPointsSubqueryPerTimeseries` limit for [subqueries](https://docs.victoriametrics.com/victoriametrics/metricsql/#subqueries). Previously it could cause runtime panic. See [#​11862](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/11862).
- BUGFIX: [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmselect` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): correctly cancel query request if `-search.skipSlowReplicas` flag is set. See [#​11771](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11771) for details.
- BUGFIX: [vmui](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#vmui): fixes predefined dashboard rows not expanding when clicking the row title. See [#​11829](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/11829). Thanks to [@​nikshrma](https://github.com/nikshrma) for contribution.
- BUGFIX: [vmui](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#vmui): support label names with special characters in series aliases. See [#​11822](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/11822). Thanks to [@​kema-dev](https://github.com/kema-dev) for contribution.
</details>
---
### Configuration
📅 **Schedule**: (UTC)
- Branch creation
- At any time (no schedule defined)
- Automerge
- At any time (no schedule defined)
🚦 **Automerge**: Disabled by config. Please merge this manually once you are satisfied.
♻ **Rebasing**: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.
🔕 **Ignore**: Close this PR and you won't be reminded about this update again.
---
- [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box
---
This PR has been generated by [Mend Renovate CLI](https://github.com/renovatebot/renovate).
<!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC4yMjUuMSIsInVwZGF0ZWRJblZlciI6IjQ0LjIyNy40IiwidGFyZ2V0QnJhbmNoIjoibWFzdGVyIiwibGFiZWxzIjpbXX0=-->
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
This PR contains the following updates:
v1.155.0→v1.156.0Release Notes
VictoriaMetrics/VictoriaMetrics (victoriametrics/vmagent)
v1.156.0Compare Source
v1.156.0
Released at 2026-11-09
FEATURE: all VictoriaMetrics Enterprise components: run with the
-eulacommand-line flag to accept ESA. License keys are no longer verified. The-license,-licenseFile,-licenseFile.reloadIntervaland-license.forceOfflineflags are deprecated and are treated as-eula, so existing setups keep working. See Running VictoriaMetrics Enterprise.FEATURE: all VictoriaMetrics components: add new metric
process_cgroup_cpu_throttled_seconds_totalwhich exposes CPU throttling. See #10430.FEATURE: vmagent and Single-node VictoriaMetrics: support
authentication_method: WorkloadIdentityin azure_sd_configs. It allows discovering Azure VMs from Kubernetes clusters with workload identity federation without a static client secret. See #11558. Thanks to @banlor for contribution.FEATURE: vmagent: add a new flag
-remoteWrite.queuePath, which allows configuring persistent-queue path for each-remoteWrite.urlindependently. See #11662.FEATURE: vmagent: add
vmagent_kafka_consumer_rebalance_events_totalandvmagent_kafka_consumer_assigned_partitionsmetrics for monitoring partition rebalancing and distribution. See partition distribution and #11370.FEATURE: vmalert: add
-notifier.sanitizeLabelNamescommand-line flag. Set it tofalseto send alert label names to notifiers without replacing characters outside[a-zA-Z0-9_:]with underscores. This is useful for Alertmanager v0.27.0 or newer, which supports UTF-8 label names when not running in classic mode. See #11629.FEATURE: vmsingle,
vmstorageandvmselectin VictoriaMetrics cluster: expose the working-set cache mode viavm_cache_modeforstorage/tsid,storage/metricIDs,storage/metricName,indexdb/tagFiltersLoops, andpromql/rollupResultcache. Values are0for split,1for switching, and2for whole. This metric helps users understand the internal memory and cache mechanisms. See #11860.BUGFIX: all VictoriaMetrics components: OS
solarisfixes compilation error. See #11850. Thanks to @szaydel for contribution.BUGFIX: all VictoriaMetrics components: correctly detect
CPUandmemorycgrouplimits configured via systemd slices incgroup v1. Previously, only the global host cgroup limit was used, which could cause OOM errors and CPU throttling. See #11827.BUGFIX: alerts: stop vmbackupmanager alerting rules
NoLatestBackupWithinLastDay,NoHourlyBackupWithinLastDay,NoDailyBackupWithinLast3Days,NoWeeklyBackupWithinLast14DaysandNoMonthlyBackupWithinLast62Daysfrom firing when scheduled backups are disabled via-disableScheduledBackupsor the corresponding backup type is disabled via-disableHourly,-disableDaily,-disableWeeklyor-disableMonthly. See #11543.BUGFIX: stream aggregation: properly delay flushes after a sudden ingestion lag spike when aggregation windows are enabled. Previously, a data gap or a restart could reduce the tolerance for delayed samples to the aggregation interval, resulting in more dropped samples than necessary. See #11866.
BUGFIX: stream aggregation: prevent skipping aggregation and deduplication flushes when aggregation windows are enabled and incoming samples are delayed by one interval or more, e.g. when the ingestion pipeline is overloaded. Previously, the flush delay could grow with the sample lag, causing skipped flushes and gaps in the results. Now the flush delay is capped at 90% of the aggregation or deduplication interval. See #11811.
BUGFIX: stream aggregation: prevent spikes in
rate_sumandrate_avgoutputs withenable_windows: truewhen out-of-order samples arrive around the aggregation window boundary. Previously, an out-of-order sample could be treated as a counter reset, causing large spike in the result. See #11810.BUGFIX: vmalert: send staleness markers for recording rule series with their original labels when a label value contains
,or=, and avoid reporting distinct label sets as duplicates when these symbols occur in label values. Previously, staleness markers could be sent with broken labels and distinct label sets could collide in the series key. See #11851. Thanks to @sarankumar1325 for contribution.BUGFIX: vmsingle and
vmselectin VictoriaMetrics cluster: properly calculate running_avg and range_avg for series withNaNpoints in the middle. Previously, everyNaNpoint was counted in the divisor, so each value after the gap moved the result by less than it should. See #11823. Thanks to @spicysharkco for contribution.BUGFIX: vmsingle and
vmselectin VictoriaMetrics cluster: do not increasevm_http_request_errors_totalwhen a query is cancelled. This prevents falseRequestErrorsToAPIalerts from firing. See #11828.BUGFIX: vmsingle and
vmselectin VictoriaMetrics cluster: properly parsemetricsqlexpression numbers with leading zeroes, such as09or08. See #11621. Thanks to @andriibeee for contribution.BUGFIX: vmsingle and
vmselectin VictoriaMetrics cluster: properly check-search.maxPointsSubqueryPerTimeserieslimit for subqueries. Previously it could cause runtime panic. See #11862.BUGFIX: vmsingle and
vmselectin VictoriaMetrics cluster: correctly cancel query request if-search.skipSlowReplicasflag is set. See #11771 for details.BUGFIX: vmui: fixes predefined dashboard rows not expanding when clicking the row title. See #11829. Thanks to @nikshrma for contribution.
BUGFIX: vmui: support label names with special characters in series aliases. See #11822. Thanks to @kema-dev for contribution.
Configuration
📅 Schedule: (UTC)
🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.
♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.
🔕 Ignore: Close this PR and you won't be reminded about this update again.
This PR has been generated by Mend Renovate CLI.