lib/storage: make lrucache ttl configurable

Previously, the `lrucache` ttl was hardcoded to 3 minutes and the code comment said it was enough for most queries that are sent to vmstorage repeatedly. But it appears that it is not always the case. `indexDB` has this `tagFilters loop cache` that is used in index search optimizations (see `getMetricIDsForDateAndFilters()`). Until recently, this cache was implemented with `workingsetcache`. In [10154](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10154), this implementation was changed to `lrucache`. After this change, some users reported huge CPU utilization increase in vmstorage (see [10297](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10297)). `workingsetcache` evicts an entry after two rotations that happen every 30 minutes. I.e. the entry ttl is 1h since the last time it was retrieved from the cache. Hence the assumption that `lrucache` causes the CPU to raise because of its too aggressive eviction policy. Reverting back to `workingsetcache` helped but `lrucache` is still preferred because it occupies less memory. Since `lrucache` is preferred, its ttl needs to be increased. Instead changing the hardcoded value, the ttl was made configurable. The `tagFilters loops cache` was configured to have ttl of 1h to match the default behavior of `workingsetcache`. Although no one complained about this yet, the ttl was also increased for `tfssCache` because previously it was also implemented with `workingsetcache`. Finally, the `regexpCache` and `prefixesCache` were also configured to have 1h ttl because these caches are also during the data retrieval. Signed-off-by: Artem Fetishev <rtm@victoriametrics.com>
docs/changelog: add known issue to v1.132.0 release notes
2026-06-21 01:28:59 +03:00 · 2026-02-09 12:01:24 +01:00 · 2026-02-09 11:03:19 +01:00 · 2026-02-06 13:17:14 +01:00 · 2026-02-05 23:52:08 +01:00 · 2026-02-05 19:48:54 +01:00
28 changed files with 157 additions and 84 deletions
--- a/app/vminsert/common/insert_ctx.go
+++ b/app/vminsert/common/insert_ctx.go
@@ -182,6 +182,7 @@ func (ctx *InsertCtx) WriteMetadata(mmpbs []prompb.MetricMetadata) error {
 		mm.Type = mmpb.Type
 		mm.Unit = bytesutil.ToUnsafeBytes(mmpb.Unit)
 	}
+	ctx.mms = mms

 	err := vmstorage.AddMetadataRows(mms)
 	if err != nil {
@@ -206,6 +207,7 @@ func (ctx *InsertCtx) WritePromMetadata(mmps []prometheus.Metadata) error {
 		mm.Help = bytesutil.ToUnsafeBytes(mmpb.Help)
 		mm.Type = mmpb.Type
 	}
+	ctx.mms = mms

 	err := vmstorage.AddMetadataRows(mms)
 	if err != nil {
--- a/app/vmui/Dockerfile-web
+++ b/app/vmui/Dockerfile-web
@@ -1,4 +1,4 @@
-FROM golang:1.25.6 AS build-web-stage
+FROM golang:1.25.7 AS build-web-stage
 COPY build /build

 WORKDIR /build
--- a/deployment/docker/Makefile
+++ b/deployment/docker/Makefile
@@ -3,11 +3,11 @@
 DOCKER_REGISTRIES ?= docker.io quay.io
 DOCKER_NAMESPACE ?= victoriametrics

-ROOT_IMAGE ?= alpine:3.23.2
+ROOT_IMAGE ?= alpine:3.23.3
 ROOT_IMAGE_SCRATCH ?= scratch
-CERTS_IMAGE := alpine:3.23.2
+CERTS_IMAGE := alpine:3.23.3

-GO_BUILDER_IMAGE := golang:1.25.6
+GO_BUILDER_IMAGE := golang:1.25.7

 BUILDER_IMAGE := local/builder:2.0.0-$(shell echo $(GO_BUILDER_IMAGE) | tr :/ __)-1
 BASE_IMAGE := local/base:1.1.4-$(shell echo $(ROOT_IMAGE) | tr :/ __)-$(shell echo $(CERTS_IMAGE) | tr :/ __)
--- a/deployment/docker/compose-vm-cluster.yml
+++ b/deployment/docker/compose-vm-cluster.yml
@@ -3,7 +3,7 @@ services:
  #  It scrapes targets defined in --promscrape.config
  #  And forward them to --remoteWrite.url
  vmagent:
-    image: victoriametrics/vmagent:v1.134.0
+    image: victoriametrics/vmagent:v1.135.0
    depends_on:
      - "vmauth"
    ports:
@@ -37,14 +37,14 @@ services:
  # vmstorage shards. Each shard receives 1/N of all metrics sent to vminserts,
  # where N is number of vmstorages (2 in this case).
  vmstorage-1:
-    image: victoriametrics/vmstorage:v1.134.0-cluster
+    image: victoriametrics/vmstorage:v1.135.0-cluster
    volumes:
      - strgdata-1:/storage
    command:
      - "--storageDataPath=/storage"
    restart: always
  vmstorage-2:
-    image: victoriametrics/vmstorage:v1.134.0-cluster
+    image: victoriametrics/vmstorage:v1.135.0-cluster
    volumes:
      - strgdata-2:/storage
    command:
@@ -54,7 +54,7 @@ services:
  # vminsert is ingestion frontend. It receives metrics pushed by vmagent,
  # pre-process them and distributes across configured vmstorage shards.
  vminsert-1:
-    image: victoriametrics/vminsert:v1.134.0-cluster
+    image: victoriametrics/vminsert:v1.135.0-cluster
    depends_on:
      - "vmstorage-1"
      - "vmstorage-2"
@@ -63,7 +63,7 @@ services:
      - "--storageNode=vmstorage-2:8400"
    restart: always
  vminsert-2:
-    image: victoriametrics/vminsert:v1.134.0-cluster
+    image: victoriametrics/vminsert:v1.135.0-cluster
    depends_on:
      - "vmstorage-1"
      - "vmstorage-2"
@@ -75,7 +75,7 @@ services:
  # vmselect is a query fronted. It serves read queries in MetricsQL or PromQL.
  # vmselect collects results from configured `--storageNode` shards.
  vmselect-1:
-    image: victoriametrics/vmselect:v1.134.0-cluster
+    image: victoriametrics/vmselect:v1.135.0-cluster
    depends_on:
      - "vmstorage-1"
      - "vmstorage-2"
@@ -85,7 +85,7 @@ services:
      - "--vmalert.proxyURL=http://vmalert:8880"
    restart: always
  vmselect-2:
-    image: victoriametrics/vmselect:v1.134.0-cluster
+    image: victoriametrics/vmselect:v1.135.0-cluster
    depends_on:
      - "vmstorage-1"
      - "vmstorage-2"
@@ -100,7 +100,7 @@ services:
  # read requests from Grafana, vmui, vmalert among vmselects.
  # It can be used as an authentication proxy.
  vmauth:
-    image: victoriametrics/vmauth:v1.134.0
+    image: victoriametrics/vmauth:v1.135.0
    depends_on:
      - "vmselect-1"
      - "vmselect-2"
@@ -114,7 +114,7 @@ services:

  # vmalert executes alerting and recording rules
  vmalert:
-    image: victoriametrics/vmalert:v1.134.0
+    image: victoriametrics/vmalert:v1.135.0
    depends_on:
      - "vmauth"
    ports:
--- a/deployment/docker/compose-vm-single.yml
+++ b/deployment/docker/compose-vm-single.yml
@@ -3,7 +3,7 @@ services:
  #  It scrapes targets defined in --promscrape.config
  #  And forward them to --remoteWrite.url
  vmagent:
-    image: victoriametrics/vmagent:v1.134.0
+    image: victoriametrics/vmagent:v1.135.0
    depends_on:
      - "victoriametrics"
    ports:
@@ -18,7 +18,7 @@ services:
  # VictoriaMetrics instance, a single process responsible for
  # storing metrics and serve read requests.
  victoriametrics:
-    image: victoriametrics/victoria-metrics:v1.134.0
+    image: victoriametrics/victoria-metrics:v1.135.0
    ports:
      - 8428:8428
      - 8089:8089
@@ -54,7 +54,7 @@ services:

  # vmalert executes alerting and recording rules
  vmalert:
-    image: victoriametrics/vmalert:v1.134.0
+    image: victoriametrics/vmalert:v1.135.0
    depends_on:
      - "victoriametrics"
      - "alertmanager"
--- a/deployment/docker/vmanomaly/vmanomaly-integration/compose.yml
+++ b/deployment/docker/vmanomaly/vmanomaly-integration/compose.yml
@@ -1,6 +1,6 @@
 services:
  vmagent:
-    image: victoriametrics/vmagent:v1.134.0
+    image: victoriametrics/vmagent:v1.135.0
    depends_on:
      - "victoriametrics"
    ports:
@@ -14,7 +14,7 @@ services:
    restart: always

  victoriametrics:
-    image: victoriametrics/victoria-metrics:v1.134.0
+    image: victoriametrics/victoria-metrics:v1.135.0
    ports:
      - 8428:8428
    volumes:
@@ -40,7 +40,7 @@ services:
    restart: always

  vmalert:
-    image: victoriametrics/vmalert:v1.134.0
+    image: victoriametrics/vmalert:v1.135.0
    depends_on:
      - "victoriametrics"
    ports:
--- a/docs/anomaly-detection/guides/guide-vmanomaly-vmalert/README.md
+++ b/docs/anomaly-detection/guides/guide-vmanomaly-vmalert/README.md
@@ -10,9 +10,9 @@ sitemap:

 - To use *vmanomaly*, part of the enterprise package, a license key is required. Obtain your key [here](https://victoriametrics.com/products/enterprise/trial/) for this tutorial or for enterprise use.
 - In the tutorial, we'll be using the following VictoriaMetrics components:
-  -  [VictoriaMetrics Single-Node](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) (v1.134.0)
-  -  [vmalert](https://docs.victoriametrics.com/victoriametrics/vmalert/) (v1.134.0)
-  -  [vmagent](https://docs.victoriametrics.com/victoriametrics/vmagent/) (v1.134.0)
+  -  [VictoriaMetrics Single-Node](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) (v1.135.0)
+  -  [vmalert](https://docs.victoriametrics.com/victoriametrics/vmalert/) (v1.135.0)
+  -  [vmagent](https://docs.victoriametrics.com/victoriametrics/vmagent/) (v1.135.0)
 - [Grafana](https://grafana.com/) (v.10.2.1)
 - [Docker](https://docs.docker.com/get-docker/) and [Docker Compose](https://docs.docker.com/compose/)
 - [Node exporter](https://github.com/prometheus/node_exporter#node-exporter) (v1.7.0) and [Alertmanager](https://prometheus.io/docs/alerting/latest/alertmanager/) (v0.27.0)
@@ -323,7 +323,7 @@ Let's wrap it all up together into the `docker-compose.yml` file.
 services:
  vmagent:
    container_name: vmagent
-    image: victoriametrics/vmagent:v1.134.0
+    image: victoriametrics/vmagent:v1.135.0
    depends_on:
      - "victoriametrics"
    ports:
@@ -340,7 +340,7 @@ services:

  victoriametrics:
    container_name: victoriametrics
-    image: victoriametrics/victoria-metrics:v1.134.0
+    image: victoriametrics/victoria-metrics:v1.135.0
    ports:
      - 8428:8428
    volumes:
@@ -373,7 +373,7 @@ services:

  vmalert:
    container_name: vmalert
-    image: victoriametrics/vmalert:v1.134.0
+    image: victoriametrics/vmalert:v1.135.0
    depends_on:
      - "victoriametrics"
    ports:
--- a/docs/guides/grafana-vmgateway-openid-configuration/README.md
+++ b/docs/guides/grafana-vmgateway-openid-configuration/README.md
@@ -249,27 +249,27 @@ services:
      - grafana_data:/var/lib/grafana/

  vmsingle:
-    image: victoriametrics/victoria-metrics:v1.134.0
+    image: victoriametrics/victoria-metrics:v1.135.0
    command:
      - -httpListenAddr=0.0.0.0:8429

  vmstorage:
-    image: victoriametrics/vmstorage:v1.134.0-cluster
+    image: victoriametrics/vmstorage:v1.135.0-cluster

  vminsert:
-    image: victoriametrics/vminsert:v1.134.0-cluster
+    image: victoriametrics/vminsert:v1.135.0-cluster
    command:
      - -storageNode=vmstorage:8400
      - -httpListenAddr=0.0.0.0:8480

  vmselect:
-    image: victoriametrics/vmselect:v1.134.0-cluster
+    image: victoriametrics/vmselect:v1.135.0-cluster
    command:
      - -storageNode=vmstorage:8401
      - -httpListenAddr=0.0.0.0:8481

  vmagent:
-    image: victoriametrics/vmagent:v1.134.0
+    image: victoriametrics/vmagent:v1.135.0
    volumes:
      - ./scrape.yaml:/etc/vmagent/config.yaml
    command:
@@ -278,7 +278,7 @@ services:
      - -remoteWrite.url=http://vmsingle:8429/api/v1/write

  vmgateway-cluster:
-    image: victoriametrics/vmgateway:v1.134.0-enterprise
+    image: victoriametrics/vmgateway:v1.135.0-enterprise
    ports:
      - 8431:8431
    volumes:
@@ -294,7 +294,7 @@ services:
      - -auth.oidcDiscoveryEndpoints=http://keycloak:8080/realms/master/.well-known/openid-configuration

  vmgateway-single:
-    image: victoriametrics/vmgateway:v1.134.0-enterprise
+    image: victoriametrics/vmgateway:v1.135.0-enterprise
    ports:
      - 8432:8431
    volumes:
@@ -405,7 +405,7 @@ Once iDP configuration is done, vmagent configuration needs to be updated to use

 ```yaml
  vmagent:
-    image: victoriametrics/vmagent:v1.134.0
+    image: victoriametrics/vmagent:v1.135.0
    volumes:
      - ./scrape.yaml:/etc/vmagent/config.yaml
      - ./vmagent-client-secret:/etc/vmagent/oauth2-client-secret
--- a/docs/victoriametrics/FAQ.md
+++ b/docs/victoriametrics/FAQ.md
@@ -286,7 +286,8 @@ Source code for Victoriametrics can be found in the following locations:
 VictoriaMetrics is able to handle data from hundreds of millions of IoT sensors and industrial sensors.
 It supports [high cardinality data](https://medium.com/@valyala/high-cardinality-tsdb-benchmarks-victoriametrics-vs-timescaledb-vs-influxdb-13e6ee64dd6b),
 perfectly [scales up on a single node](https://medium.com/@valyala/measuring-vertical-scalability-for-time-series-databases-in-google-cloud-92550d78d8ae)
-and scales horizontally to multiple nodes.
+and scales horizontally to multiple nodes in [cluster setup](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/).
+It also supports an option for reducing the index size for IoT data - see [these docs](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#index-tuning-for-low-churn-rate).

 ## What is the difference between single-node and cluster versions of VictoriaMetrics?

--- a/docs/victoriametrics/Quick-Start.md
+++ b/docs/victoriametrics/Quick-Start.md
@@ -58,9 +58,9 @@ Download the newest available [VictoriaMetrics release](https://docs.victoriamet
 from [DockerHub](https://hub.docker.com/r/victoriametrics/victoria-metrics) or [Quay](https://quay.io/repository/victoriametrics/victoria-metrics?tab=tags):

 ```sh
-docker pull victoriametrics/victoria-metrics:v1.134.0
+docker pull victoriametrics/victoria-metrics:v1.135.0
 docker run -it --rm -v `pwd`/victoria-metrics-data:/victoria-metrics-data -p 8428:8428 \
- victoriametrics/victoria-metrics:v1.134.0 --selfScrapeInterval=5s -storageDataPath=victoria-metrics-data
+ victoriametrics/victoria-metrics:v1.135.0 --selfScrapeInterval=5s -storageDataPath=victoria-metrics-data
 ```

 _For Enterprise images see [this link](https://docs.victoriametrics.com/victoriametrics/enterprise/#docker-images)._
--- a/docs/victoriametrics/README.md
+++ b/docs/victoriametrics/README.md
@@ -93,8 +93,10 @@ VictoriaMetrics has the following prominent features:
 * It supports metrics [relabeling](#relabeling).
 * It can deal with [high cardinality issues](https://docs.victoriametrics.com/victoriametrics/faq/#what-is-high-cardinality) and
  [high churn rate](https://docs.victoriametrics.com/victoriametrics/faq/#what-is-high-churn-rate) issues via [series limiter](#cardinality-limiter).
-* It ideally works with big amounts of time series data from APM, Kubernetes, IoT sensors, connected cars, industrial telemetry, financial data
-  and various [Enterprise workloads](https://docs.victoriametrics.com/victoriametrics/enterprise/).
+* It ideally works for big amounts of time series with both [high churn rate](https://docs.victoriametrics.com/victoriametrics/faq/#what-is-high-churn-rate) (APM, Kubernetes)
+  and low churn rate (IoT sensors, connected cars, industrial telemetry, financial data - see
+  [these docs](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#index-tuning-for-low-churn-rate)),
+  plus various [Enterprise workloads](https://docs.victoriametrics.com/victoriametrics/enterprise/).
 * It has an open source [cluster version](https://github.com/VictoriaMetrics/VictoriaMetrics/tree/cluster).
 * It can store data on [NFS-based storages](https://en.wikipedia.org/wiki/Network_File_System) such as [Amazon EFS](https://aws.amazon.com/efs/)
  and [Google Filestore](https://cloud.google.com/filestore).
@@ -1671,6 +1673,9 @@ See [Why IndexDB size is so large?](https://docs.victoriametrics.com/victoriamet
 Downsampling is performed during [background merges](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#storage).
 It cannot be performed if there is not enough of free disk space or if vmstorage is in [read-only mode](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/#readonly-mode).

+Downsampling period changes `/api/v1/export` API output. During query requests, if export `start` period is not specified and `reduce_mem_usage` param is omitted, the biggest `downsampling.period` is applied.
+As an example, export request  `/api/v1/export?match[]=series` with `-downsampling.period=30d:1h,180d:24h` will return samples downsampled with `24h` interval.
+
 It's expected that resource usage will temporarily increase when **downsampling with filters** is applied.
 This is because additional operations are required to read historical data, downsample, and persist it back,
 which will cost extra CPU and memory.
--- a/docs/victoriametrics/changelog/CHANGELOG.md
+++ b/docs/victoriametrics/changelog/CHANGELOG.md
@@ -25,6 +25,13 @@ The sandbox cluster installation runs under the constant load generated by
 See also [LTS releases](https://docs.victoriametrics.com/victoriametrics/lts-releases/).

 ## tip
+* SECURITY: upgrade Go builder from Go1.25.6 to Go1.25.7. See [the list of issues addressed in Go1.25.7](https://github.com/golang/go/issues?q=milestone%3AGo1.25.7%20label%3ACherryPickApproved).
+* SECURITY: upgrade base docker image (Alpine) from 3.23.2 to 3.23.3. See [Alpine 3.23.3 release notes](https://www.alpinelinux.org/posts/Alpine-3.20.9-3.21.6-3.22.3-3.23.3-released.html).
+
+* BUGFIX: all VictoriaMetrics components: respect default http client proxy env variables (HTTP_PROXY,HTTPS_PROXY,NO_PROXY). See [#10385](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10385). Thanks to @zane-deg for the contribution.
+* BUGFIX: [vmagent](https://docs.victoriametrics.com/vmagent/) and [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/): properly expose `kubernetes_sd` discovery network dialer metrics `vm_promscrape_discovery_kubernetes_conn_*`. See [#10382](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10382).
+* BUGFIX: [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/): slightly reduce memory usage for [metrics-metadata](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#metrics-metadata) ingestion. See [#10392](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10392).
+* BUGFIX: [vmagent](https://docs.victoriametrics.com/vmagent/) and [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/): introduce time‑based manual offset commit for kafka consumer to fix performance degradation with enabled manual commit. After this change, it will commit partition offsets in batch per second to avoid high commit QPS on the Kafka broker. It's no longer recommended to set `enable.auto.commit=true` in `-kafka.consumer.topic.options`, as `vmagent` will automatically manage it. See [#10395](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10395).

 ## [v1.135.0](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.135.0)

@@ -84,7 +91,7 @@ Released at 2026-01-16
 Released at 2026-01-02

 **Update Note 1:** [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): Upgrading to per-partition index requires registering all active time series. Expect slow down of data ingestion and queries during upgrade roll-out. This is a one-time operation. Additionally, for users with retention periods shorter than 1 month the disk usage may increase.
-**Update Note 2:** [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): Running this version in deployments with big datasets may cause high CPU utilization. See [10297](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10297).
+**Update Note 2:** [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): Certain data and query patterns may cause high CPU utilization due to [10154](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10154). See [10297](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10297).
 **Update Note 3:** [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): lock contention in the ingestion path may cause frequent context switches and storage connection saturation spikes. See [10367](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10367). Addressed in `v1.135.0`.
 **Update Note 4:** [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): TSDB status may be empty if the partition index does not have records for the requested date. See [10315](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10315). Addressed in `v1.135.0`.
 **Update Note 5:** [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): `indexdb/tagFiltersToMetricIDs`, `indexdb/metricID` and `indexdb/date_metricID` cache metrics are not reported properly. See [10275](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10275). Addressed in `v1.135.0`.
@@ -153,6 +160,22 @@ See changes [here](https://docs.victoriametrics.com/victoriametrics/changelog/ch

 See changes [here](https://docs.victoriametrics.com/victoriametrics/changelog/changelog_2025/#v11230)

+## [v1.122.14](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.122.14)
+
+Released at 2026-01-30
+
+**v1.122.x is a line of [LTS releases](https://docs.victoriametrics.com/victoriametrics/lts-releases/). It contains important up-to-date bugfixes for [VictoriaMetrics enterprise](https://docs.victoriametrics.com/victoriametrics/enterprise/).
+All these fixes are also included in [the latest community release](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/latest).
+The v1.122.x line will be supported for at least 12 months since [v1.122.0](https://docs.victoriametrics.com/victoriametrics/changelog/#v11220) release**
+
+* BUGFIX: [vmauth](https://docs.victoriametrics.com/victoriametrics/vmauth/): stop backend health checks for URL prefixes defined in `url_map` during configuration reloads. Previously, stale backends kept being health-checked and produced repeated warning logs after reloads. See [#10334](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10334).
+* BUGFIX: `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): correctly return tenants results for `/admin/tenants` when `start` or `end` are specified. See [#10312](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10312)
+* BUGFIX: [vmui](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#vmui): fix "Percentage from total" calculation on the Cardinality Explorer page when multiple metrics match the filter. See [#10323](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10323). Thanks to @PleasingFungus for the contribution.
+* BUGFIX: [vmagent](https://docs.victoriametrics.com/victoriametrics/vmagent/): apply `-promscrape.maxScrapeSize` check to decompressed data instead of compressed data. See [#9481](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/9481).
+* BUGFIX: [vmalert](https://docs.victoriametrics.com/victoriametrics/vmalert/): fix [alert restore](https://docs.victoriametrics.com/victoriametrics/vmalert/#alerts-state-on-restarts) when a group contains many rules and is slow to complete evaluation. Previously, the restore process might not retrieve the correct previous alert state. See [#10335](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10335).
+* BUGFIX: [MetricsQL](https://docs.victoriametrics.com/victoriametrics/metricsql/): fix `changes()` function when gaps between samples exceed the lookbehind window. Previously, it could yield a non-zero value even when the sample value remained unchanged. See [#10280](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10280).
+* BUGFIX: [vmauth](https://docs.victoriametrics.com/victoriametrics/vmauth/): fix an issue where canceling a client request (closing a browser tab or timeout) incorrectly marked all backends as unavailable for `-failTimeout` duration (3s by default), even though backends were healthy. See [#10318](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10318).
+
 ## [v1.122.13](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.122.13)

 Released at 2026-01-16
@@ -268,6 +291,21 @@ See changes [here](https://docs.victoriametrics.com/victoriametrics/changelog/ch

 See changes [here](https://docs.victoriametrics.com/victoriametrics/changelog/changelog_2025/#v11110)

+## [v1.110.29](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.110.29)
+
+Released at 2026-01-30
+
+**v1.110.x is a line of [LTS releases](https://docs.victoriametrics.com/victoriametrics/lts-releases/). It contains important up-to-date bugfixes for [VictoriaMetrics enterprise](https://docs.victoriametrics.com/victoriametrics/enterprise/).
+All these fixes are also included in [the latest community release](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/latest).
+The v1.110.x line will be supported for at least 12 months since [v1.110.0](https://docs.victoriametrics.com/victoriametrics/changelog/#v11100) release**
+
+* BUGFIX: [MetricsQL](https://docs.victoriametrics.com/victoriametrics/metricsql/): fix `changes()` function when gaps between samples exceed the lookbehind window. Previously, it could yield a non-zero value even when the sample value remained unchanged. See [#10280](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10280).
+* BUGFIX: [vmalert](https://docs.victoriametrics.com/victoriametrics/vmalert/): fix [alert restore](https://docs.victoriametrics.com/victoriametrics/vmalert/#alerts-state-on-restarts) when a group contains many rules and is slow to complete evaluation. Previously, the restore process might not retrieve the correct previous alert state. See [#10335](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10335).
+* BUGFIX: [vmauth](https://docs.victoriametrics.com/victoriametrics/vmauth/): fix an issue where canceling a client request (closing a browser tab or timeout) incorrectly marked all backends as unavailable for `-failTimeout` duration (3s by default), even though backends were healthy. See [#10318](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10318).
+* BUGFIX: [vmauth](https://docs.victoriametrics.com/victoriametrics/vmauth/): stop backend health checks for URL prefixes defined in `url_map` during configuration reloads. Previously, stale backends kept being health-checked and produced repeated warning logs after reloads. See [#10334](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10334).
+* BUGFIX: `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): correctly return tenants results for `/admin/tenants` when `start` or `end` are specified. See [#10312](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10312)
+* BUGFIX: [vmui](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#vmui): fix "Percentage from total" calculation on the Cardinality Explorer page when multiple metrics match the filter. See [#10323](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10323). Thanks to @PleasingFungus for the contribution.
+
 ## [v1.110.28](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.110.28)

 Released at 2026-01-16
--- a/docs/victoriametrics/changelog/CHANGELOG_2025.md
+++ b/docs/victoriametrics/changelog/CHANGELOG_2025.md
@@ -78,6 +78,8 @@ Read [VictoriaLogs#869](https://github.com/VictoriaMetrics/VictoriaLogs/issues/8

 **Known issue: `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): may shutdown ungracefully (data, indexes, and caches are not flushed to disk) in Kubernetes deployments if the number of `vminserts` is > 1 and `terminationGracePeriodSeconds` < 60s (30s by default). The issue was introduced in [9487](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/9487) and [10136](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10136) and was fixed in [10224](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10224). The fix will be included into `v1.133.0`. Even with this fix, ungraceful shutdowns are possible if flushing in-memory contents takes more than 5s. In this case decrease `-storage.vminsertConnsShutdownDuration` and/or increase [terminationGracePeriodSeconds](https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle/#pod-termination-flow).**

+**Known issue:** [vmsingle](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) and `vmstorage` in [VictoriaMetrics cluster](https://docs.victoriametrics.com/victoriametrics/cluster-victoriametrics/): Certain data and query patterns may cause high CPU utilization due to [10154](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10154). See [10297](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10297).
+
 * SECURITY: upgrade Go builder from Go1.25.4 to Go1.25.5. See [the list of issues addressed in Go1.25.5](https://github.com/golang/go/issues?q=milestone%3AGo1.25.5%20label%3ACherryPickApproved).

 * FEATURE: [dashboards/operator](https://grafana.com/grafana/dashboards/17869-victoriametrics-operator/): add panels for flags and configuration parameters values. See [operator#1341](https://github.com/VictoriaMetrics/operator/issues/1341).
@@ -2257,4 +2259,4 @@ See changes [here](https://docs.victoriametrics.com/victoriametrics/changelog/ch

 ## Previous releases

-See changes for older releases [here](https://docs.victoriametrics.com/victoriametrics/changelog/changelog_2024/).
+See changes for older releases [here](https://docs.victoriametrics.com/victoriametrics/changelog/changelog_2024/).
--- a/docs/victoriametrics/enterprise.md
+++ b/docs/victoriametrics/enterprise.md
@@ -117,7 +117,7 @@ It is allowed to run VictoriaMetrics and VictoriaLogs Enterprise components in [

 Binary releases of Enterprise components are available at [the releases page for VictoriaMetrics](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/latest)
 and [the releases page for VictoriaLogs](https://github.com/VictoriaMetrics/VictoriaLogs/releases/latest).
-Enterprise binaries and packages have `enterprise` suffix in their names. For example, `victoria-metrics-linux-amd64-v1.134.0-enterprise.tar.gz`.
+Enterprise binaries and packages have `enterprise` suffix in their names. For example, `victoria-metrics-linux-amd64-v1.135.0-enterprise.tar.gz`.

 In order to run binary release of Enterprise component, please download the `*-enterprise.tar.gz` archive for your OS and architecture
 from the corresponding releases page and unpack it. Then run the unpacked binary.
@@ -135,8 +135,8 @@ For example, the following command runs VictoriaMetrics Enterprise binary with t
 obtained at [this page](https://victoriametrics.com/products/enterprise/trial/):

 ```sh
-wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.134.0/victoria-metrics-linux-amd64-v1.134.0-enterprise.tar.gz
-tar -xzf victoria-metrics-linux-amd64-v1.134.0-enterprise.tar.gz
+wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.135.0/victoria-metrics-linux-amd64-v1.135.0-enterprise.tar.gz
+tar -xzf victoria-metrics-linux-amd64-v1.135.0-enterprise.tar.gz
 ./victoria-metrics-prod -license=BASE64_ENCODED_LICENSE_KEY
 ```

@@ -151,7 +151,7 @@ Alternatively, VictoriaMetrics Enterprise license can be stored in the file and
 It is allowed to run VictoriaMetrics and VictoriaLogs Enterprise components in [cases listed here](#valid-cases-for-victoriametrics-enterprise).

 Docker images for Enterprise components are available at [VictoriaMetrics Docker Hub](https://hub.docker.com/u/victoriametrics) and [VictoriaMetrics Quay](https://quay.io/organization/victoriametrics).
-Enterprise docker images have `enterprise` suffix in their names. For example, `victoriametrics/victoria-metrics:v1.134.0-enterprise`.
+Enterprise docker images have `enterprise` suffix in their names. For example, `victoriametrics/victoria-metrics:v1.135.0-enterprise`.

 In order to run Docker image of VictoriaMetrics Enterprise component, it is required to provide the license key via the command-line
 flag as described in the [binary-releases](#binary-releases) section.
@@ -161,13 +161,13 @@ Enterprise license key can be obtained at [this page](https://victoriametrics.co
 For example, the following command runs VictoriaMetrics Enterprise Docker image with the specified license key:

 ```sh
-docker run --name=victoria-metrics victoriametrics/victoria-metrics:v1.134.0-enterprise -license=BASE64_ENCODED_LICENSE_KEY
+docker run --name=victoria-metrics victoriametrics/victoria-metrics:v1.135.0-enterprise -license=BASE64_ENCODED_LICENSE_KEY
 ```

 Alternatively, the license code can be stored in the file and then referred via `-licenseFile` command-line flag:

 ```sh
-docker run --name=victoria-metrics -v /vm-license:/vm-license  victoriametrics/victoria-metrics:v1.134.0-enterprise -licenseFile=/path/to/vm-license
+docker run --name=victoria-metrics -v /vm-license:/vm-license  victoriametrics/victoria-metrics:v1.135.0-enterprise -licenseFile=/path/to/vm-license
 ```

 Example docker-compose configuration:
@@ -177,7 +177,7 @@ version: "3.5"
 services:
  victoriametrics:
    container_name: victoriametrics
-    image: victoriametrics/victoria-metrics:v1.134.0
+    image: victoriametrics/victoria-metrics:v1.135.0
    ports:
      - 8428:8428
    volumes:
@@ -209,7 +209,7 @@ is used to provide the license key in plain-text:
 ```yaml
 server:
  image:
-    tag: v1.134.0-enterprise
+    tag: v1.135.0-enterprise

 license:
  key: {BASE64_ENCODED_LICENSE_KEY}
@@ -220,7 +220,7 @@ In order to provide the license key via existing secret, the following values fi
 ```yaml
 server:
  image:
-    tag: v1.134.0-enterprise
+    tag: v1.135.0-enterprise

 license:
  secret:
@@ -270,7 +270,7 @@ spec:
  license:
    key: {BASE64_ENCODED_LICENSE_KEY}
  image:
-    tag: v1.134.0-enterprise
+    tag: v1.135.0-enterprise
 ```

 In order to provide the license key via an existing secret, the following custom resource is used:
@@ -287,7 +287,7 @@ spec:
      name: vm-license
      key: license
  image:
-    tag: v1.134.0-enterprise
+    tag: v1.135.0-enterprise
 ```

 Example secret with license key:
@@ -338,7 +338,7 @@ Builds are available for amd64 and arm64 architectures.

 Example archive:

-`victoria-metrics-linux-amd64-v1.134.0-enterprise.tar.gz`
+`victoria-metrics-linux-amd64-v1.135.0-enterprise.tar.gz`

 Includes:

@@ -347,7 +347,7 @@ Includes:

 Example Docker image:

-`victoriametrics/victoria-metrics:v1.134.0-enterprise-fips` – uses the FIPS-compatible binary and based on `scratch` image.
+`victoriametrics/victoria-metrics:v1.135.0-enterprise-fips` – uses the FIPS-compatible binary and based on `scratch` image.

 ## Monitoring license expiration

--- a/docs/victoriametrics/integrations/zabbixconnector.md
+++ b/docs/victoriametrics/integrations/zabbixconnector.md
@@ -5,6 +5,8 @@ menu:
  docs:
    parent: "integrations-vm"
    weight: 10
+aliases:
+  - /victoriametrics/integrations/zabbix/
 ---

 VictoriaMetrics components like **vmagent**, **vminsert** or **single-node** can receive data from 
@@ -61,4 +63,4 @@ curl http://localhost:8428/api/v1/export -d 'match={host="Zabbix server"}'
 {"metric":{"__name__":"item_1","host":"ZabbixServer","hostname":"ZabbixServer","group_servers":"exists","tag_foo":"exists"},"values":[0],"timestamps":[1673454303800]}
 {"metric":{"__name__":"item_2","host":"ZabbixServer","hostname":"ZabbixServer","group_servers":"exists","tag_foo":"test,exists"},"values":[1],"timestamps":[1673454303832]}
 {"metric":{"__name__":"item_3","host":"ZabbixServer","hostname":"ZabbixServer","group_servers":"exists","tag_bar":"test"},"values":[123],"timestamps":[1673454303867]}
-```
+```
--- a/docs/victoriametrics/scrape_config_examples.md
+++ b/docs/victoriametrics/scrape_config_examples.md
@@ -35,8 +35,8 @@ scrape_configs:
 After you created the `scrape.yaml` file, download and unpack [single-node VictoriaMetrics](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/) to the same directory:

 ```sh
-wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.134.0/victoria-metrics-linux-amd64-v1.134.0.tar.gz
-tar xzf victoria-metrics-linux-amd64-v1.134.0.tar.gz
+wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.135.0/victoria-metrics-linux-amd64-v1.135.0.tar.gz
+tar xzf victoria-metrics-linux-amd64-v1.135.0.tar.gz
 ```

 Then start VictoriaMetrics and instruct it to scrape targets defined in `scrape.yaml` and save scraped metrics
@@ -150,8 +150,8 @@ Then start [single-node VictoriaMetrics](https://docs.victoriametrics.com/victor

 ```yaml
 # Download and unpack single-node VictoriaMetrics
-wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.134.0/victoria-metrics-linux-amd64-v1.134.0.tar.gz
-tar xzf victoria-metrics-linux-amd64-v1.134.0.tar.gz
+wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.135.0/victoria-metrics-linux-amd64-v1.135.0.tar.gz
+tar xzf victoria-metrics-linux-amd64-v1.135.0.tar.gz

 # Run single-node VictoriaMetrics with the given scrape.yaml
 ./victoria-metrics-prod -promscrape.config=scrape.yaml
--- a/docs/victoriametrics/vmagent.md
+++ b/docs/victoriametrics/vmagent.md
@@ -114,6 +114,8 @@ The maximum on-disk size for the buffered metrics can be limited with `-remoteWr

 `vmagent` can save on network bandwidth usage costs by using [VictoriaMetrics remote write protocol](#victoriametrics-remote-write-protocol).

+See [how to optimize index size at VictoriaMetrics for IoT and industrial monitoring](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#index-tuning-for-low-churn-rate).
+
 ### Drop-in replacement for Prometheus

 If you use Prometheus only for scraping metrics from various targets and forwarding these metrics to remote storage
--- a/docs/victoriametrics/vmctl/vmctl.md
+++ b/docs/victoriametrics/vmctl/vmctl.md
@@ -34,9 +34,9 @@ vmctl command-line tool is available as:

 Download and unpack vmctl:
 ```sh
-wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.134.0/vmutils-darwin-arm64-v1.134.0.tar.gz
+wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.135.0/vmutils-darwin-arm64-v1.135.0.tar.gz

-tar xzf vmutils-darwin-arm64-v1.134.0.tar.gz
+tar xzf vmutils-darwin-arm64-v1.135.0.tar.gz
 ```

 Once binary is unpacked, see the full list of supported modes by running the following command:
--- a/go.mod
+++ b/go.mod
@@ -1,6 +1,6 @@
 module github.com/VictoriaMetrics/VictoriaMetrics

-go 1.25.6
+go 1.25.7

 require (
 	cloud.google.com/go/storage v1.59.1
--- a/lib/lrucache/lrucache.go
+++ b/lib/lrucache/lrucache.go
@@ -28,8 +28,12 @@ type Cache struct {
 // NewCache creates new cache.
 //
 // Cache size in bytes is limited by the value returned by getMaxSizeBytes() callback.
+//
+// Entries will be removed from the cache if not accessed longer than ttl
+// duration but not sooner than ~53 seconds (see cleaner() below).
+//
 // Call MustStop() in order to free up resources occupied by Cache.
-func NewCache(getMaxSizeBytes func() uint64) *Cache {
+func NewCache(getMaxSizeBytes func() uint64, ttl time.Duration) *Cache {
 	cpusCount := cgroup.AvailableCPUs()
 	shardsCount := cgroup.AvailableCPUs()
 	// Increase the number of shards with the increased number of available CPU cores.
@@ -45,7 +49,7 @@ func NewCache(getMaxSizeBytes func() uint64) *Cache {
 		return n / uint64(shardsCount)
 	}
 	for i := range shards {
-		shards[i] = newCache(getMaxShardBytes)
+		shards[i] = newCache(getMaxShardBytes, ttl)
 	}
 	c := &Cache{
 		shards:            shards,
@@ -175,6 +179,11 @@ type cache struct {
 	// getMaxSizeBytes() is a callback, which returns the maximum allowed cache size in bytes.
 	getMaxSizeBytes func() uint64

+	// ttl is the duration in seconds during which an entry is allowed to stay
+	// in the cache without being retrieved. Once this time passes the entry
+	// will be removed from the cache.
+	ttl uint64
+
 	// mu protects all the fields below.
 	mu sync.Mutex

@@ -210,9 +219,10 @@ type cacheEntry struct {
 	e Entry
 }

-func newCache(getMaxSizeBytes func() uint64) *cache {
+func newCache(getMaxSizeBytes func() uint64, ttl time.Duration) *cache {
 	var c cache
 	c.getMaxSizeBytes = getMaxSizeBytes
+	c.ttl = uint64(ttl.Seconds())
 	c.m = make(map[string]*cacheEntry)
 	return &c
 }
@@ -231,9 +241,7 @@ func (c *cache) updateSizeBytes(n uint64) {
 }

 func (c *cache) cleanByTimeout() {
-	// Delete items accessed more than three minutes ago.
-	// This time should be enough for repeated queries.
-	lastAccessTime := fasttime.UnixTimestamp() - 3*60
+	lastAccessTime := fasttime.UnixTimestamp() - c.ttl
 	c.mu.Lock()
 	defer c.mu.Unlock()

--- a/lib/lrucache/lrucache_test.go
+++ b/lib/lrucache/lrucache_test.go
@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"sync"
 	"testing"
+	"time"

 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/cgroup"
 )
@@ -18,7 +19,7 @@ func TestCache(t *testing.T) {
 	getMaxSize := func() uint64 {
 		return sizeMaxBytes
 	}
-	c := NewCache(getMaxSize)
+	c := NewCache(getMaxSize, 3*time.Minute)
 	defer c.MustStop()
 	if n := c.SizeBytes(); n != 0 {
 		t.Fatalf("unexpected SizeBytes(); got %d; want %d", n, 0)
@@ -123,7 +124,7 @@ func TestCacheConcurrentAccess(_ *testing.T) {
 	getMaxSize := func() uint64 {
 		return sizeMaxBytes
 	}
-	c := NewCache(getMaxSize)
+	c := NewCache(getMaxSize, 3*time.Minute)
 	defer c.MustStop()

 	workers := 5
--- a/lib/promauth/config.go
+++ b/lib/promauth/config.go
@@ -234,7 +234,9 @@ func urlValuesFromMap(m map[string]string) url.Values {

 func (oi *oauth2ConfigInternal) initTokenSource() error {
 	tr := httputil.NewTransport(false, "vm_oauth_client")
-	tr.Proxy = oi.proxyURLFunc
+	if oi.proxyURLFunc != nil {
+		tr.Proxy = oi.proxyURLFunc
+	}
 	c := &http.Client{
 		Transport: oi.ac.NewRoundTripper(tr),
 	}
--- a/lib/promscrape/client.go
+++ b/lib/promscrape/client.go
@@ -81,7 +81,9 @@ func newClient(ctx context.Context, sw *ScrapeWork) (*client, error) {
 	}

 	tr := httputil.NewTransport(false, "vm_promscrape")
-	tr.Proxy = proxyURLFunc
+	if proxyURLFunc != nil {
+		tr.Proxy = proxyURLFunc
+	}
 	tr.TLSHandshakeTimeout = 10 * time.Second
 	tr.IdleConnTimeout = 2 * sw.ScrapeInterval
 	tr.DisableKeepAlives = *disableKeepAlive || sw.DisableKeepAlive
--- a/lib/promscrape/discovery/kubernetes/api_watcher.go
+++ b/lib/promscrape/discovery/kubernetes/api_watcher.go
@@ -22,7 +22,6 @@ import (
 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/cgroup"
 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/httputil"
 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/logger"
-	"github.com/VictoriaMetrics/VictoriaMetrics/lib/netutil"
 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/promauth"
 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/promutil"
 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/timerpool"
@@ -269,11 +268,9 @@ func getHTTPClient(ac *promauth.Config, proxyURL *url.URL) *http.Client {
 	if !*useHTTP2Client {
 		// Proxy is not supported for http2 client.
 		// See https://github.com/golang/go/issues/26479
-		var proxy func(*http.Request) (*url.URL, error)
 		if proxyURL != nil {
-			proxy = http.ProxyURL(proxyURL)
+			tr.Proxy = http.ProxyURL(proxyURL)
 		}
-		tr.Proxy = proxy
 	}
 	c := &http.Client{
 		Transport: ac.NewRoundTripper(tr),
@@ -286,7 +283,6 @@ func getHTTPClient(ac *promauth.Config, proxyURL *url.URL) *http.Client {

 func newHTTPTransport(enableHTTP2 bool) *http.Transport {
 	tr := httputil.NewTransport(enableHTTP2, "vm_promscrape_discovery_kubernetes")
-	tr.DialContext = netutil.Dialer.DialContext
 	tr.TLSHandshakeTimeout = 10 * time.Second
 	tr.IdleConnTimeout = *apiServerTimeout
 	tr.MaxIdleConnsPerHost = 100
--- a/lib/promscrape/discoveryutil/client.go
+++ b/lib/promscrape/discoveryutil/client.go
@@ -112,7 +112,9 @@ func NewClient(apiServer string, ac *promauth.Config, proxyURL *proxy.URL, proxy
 	}

 	tr := httputil.NewTransport(false, "vm_promscrape_discovery")
-	tr.Proxy = proxyURLFunc
+	if proxyURLFunc != nil {
+		tr.Proxy = proxyURLFunc
+	}
 	tr.TLSHandshakeTimeout = 10 * time.Second
 	tr.MaxIdleConnsPerHost = *maxConcurrency
 	tr.ResponseHeaderTimeout = DefaultClientReadTimeout
@@ -124,7 +126,9 @@ func NewClient(apiServer string, ac *promauth.Config, proxyURL *proxy.URL, proxy
 	}

 	trBlocking := httputil.NewTransport(false, "vm_promscrape_discovery")
-	trBlocking.Proxy = proxyURLFunc
+	if proxyURLFunc != nil {
+		trBlocking.Proxy = proxyURLFunc
+	}
 	trBlocking.TLSHandshakeTimeout = 10 * time.Second
 	trBlocking.MaxIdleConnsPerHost = 1000
 	trBlocking.ResponseHeaderTimeout = BlockingClientReadTimeout
--- a/lib/protoparser/protoparserutil/compress_reader.go
+++ b/lib/protoparser/protoparserutil/compress_reader.go
@@ -74,7 +74,14 @@ func ReadUncompressedData(r io.Reader, contentType string, maxDataSize *flagutil
 func readUncompressedData(r io.Reader, maxDataSize *flagutil.Bytes, decompress func(dst, src []byte) ([]byte, error), callback func(data []byte) error) error {
 	return readFull(r, maxDataSize, func(data []byte) error {
 		dbb := decompressedBufPool.Get()
-		defer decompressedBufPool.Put(dbb)
+		defer func() {
+			if cap(dbb.B) > 64*1024 && cap(dbb.B) > 4*len(dbb.B) {
+				// Do not store too big dbb to the pool if only a small part of the buffer is used last time.
+				// This should reduce memory waste.
+				return
+			}
+			decompressedBufPool.Put(dbb)
+		}()

 		var err error
 		dbb.B, err = decompress(dbb.B, data)
@@ -95,7 +102,7 @@ func readFull(r io.Reader, maxDataSize *flagutil.Bytes, callback func(data []byt

 	bb := fullReaderBufPool.Get()
 	defer func() {
-		if len(bb.B) > 1024*1024 && cap(bb.B) > 4*len(bb.B) {
+		if cap(bb.B) > 64*1024 && cap(bb.B) > 4*len(bb.B) {
 			// Do not store too big bb to the pool if only a small part of the buffer is used last time.
 			// This should reduce memory waste.
 			return
--- a/lib/storage/index_db.go
+++ b/lib/storage/index_db.go
@@ -173,7 +173,7 @@ func mustOpenIndexDB(id uint64, tr TimeRange, name, path string, s *Storage, isR
 		logger.Panicf("BUG: Storage must not be nil")
 	}

-	tfssCache := lrucache.NewCache(getTagFiltersCacheSize)
+	tfssCache := lrucache.NewCache(getTagFiltersCacheSize, 1*time.Hour)
 	tb := mergeset.MustOpenTable(path, dataFlushInterval, tfssCache.Reset, mergeTagToMetricIDsRows, isReadOnly)
 	db := &indexDB{
 		legacyMinMissingTimestampByKey: make(map[string]int64),
@@ -183,7 +183,7 @@ func mustOpenIndexDB(id uint64, tr TimeRange, name, path string, s *Storage, isR
 		tb:                             tb,
 		s:                              s,
 		tagFiltersToMetricIDsCache:     tfssCache,
-		loopsPerDateTagFilterCache:     lrucache.NewCache(getTagFiltersLoopsCacheSize),
+		loopsPerDateTagFilterCache:     lrucache.NewCache(getTagFiltersLoopsCacheSize, 1*time.Hour),
 		metricIDCache:                  newMetricIDCache(),
 		dateMetricIDCache:              newDateMetricIDCache(),
 	}
--- a/lib/storage/tag_filters.go
+++ b/lib/storage/tag_filters.go
@@ -8,6 +8,7 @@ import (
 	"sort"
 	"strings"
 	"sync"
+	"time"
 	"unsafe"

 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/atomicutil"
@@ -866,7 +867,7 @@ var (
 )

 var (
-	regexpCache = lrucache.NewCache(getMaxRegexpCacheSize)
+	regexpCache = lrucache.NewCache(getMaxRegexpCacheSize, 1*time.Hour)
 )

 type regexpCacheValue struct {
@@ -921,7 +922,7 @@ var (
 )

 var (
-	prefixesCache = lrucache.NewCache(getMaxPrefixesCacheSize)
+	prefixesCache = lrucache.NewCache(getMaxPrefixesCacheSize, 1*time.Hour)
 )

 // RegexpPrefixesCacheSize returns the number of cached regexp prefixes for tag filters.
Author	SHA1	Message	Date
Artem Fetishev	3f1b7a3a7f	lib/storage: make lrucache ttl configurable Previously, the `lrucache` ttl was hardcoded to 3 minutes and the code comment said it was enough for most queries that are sent to vmstorage repeatedly. But it appears that it is not always the case. `indexDB` has this `tagFilters loop cache` that is used in index search optimizations (see `getMetricIDsForDateAndFilters()`). Until recently, this cache was implemented with `workingsetcache`. In [10154](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10154), this implementation was changed to `lrucache`. After this change, some users reported huge CPU utilization increase in vmstorage (see [10297](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10297)). `workingsetcache` evicts an entry after two rotations that happen every 30 minutes. I.e. the entry ttl is 1h since the last time it was retrieved from the cache. Hence the assumption that `lrucache` causes the CPU to raise because of its too aggressive eviction policy. Reverting back to `workingsetcache` helped but `lrucache` is still preferred because it occupies less memory. Since `lrucache` is preferred, its ttl needs to be increased. Instead changing the hardcoded value, the ttl was made configurable. The `tagFilters loops cache` was configured to have ttl of 1h to match the default behavior of `workingsetcache`. Although no one complained about this yet, the ttl was also increased for `tfssCache` because previously it was also implemented with `workingsetcache`. Finally, the `regexpCache` and `prefixesCache` were also configured to have 1h ttl because these caches are also during the data retrieval. Signed-off-by: Artem Fetishev <rtm@victoriametrics.com>	2026-02-09 12:01:24 +01:00
Artem Fetishev	5915c57b46	docs/changelog: add known issue to v1.132.0 release notes Signed-off-by: Artem Fetishev <rtm@victoriametrics.com>	2026-02-09 11:03:19 +01:00
JAYICE	f36e1857c0	app/vmagent: improve kafka consumer performance Previously, the Kafka consumer in vmagent committed offsets per message (manual commit). At high message rates, this could overload the commit path (coordinator, __consumer_offsets topic, and network). This commit introduces time-based manual commits with a controlled window: * enable.auto.commit remains false by default. * After a successful TryPush (data accepted into the buffer before the vmagent queue/backend), vmagent adds the message to pending offsets. * Offsets are committed periodically (every second), as well as during shutdown and partition rebalance. This keeps the commit point tied to TryPush (stronger guarantees than auto-commit) while significantly reducing commit QPS. Auto-commit is also time-based, but it advances offsets based on poll() delivery rather than application-level processing. This means offsets may be committed before data is actually accepted by the vmagent pipeline, slightly increasing the risk of data loss on crash or restart. This change does not make the Kafka consumer fully transactional end-to-end. Buffers in vmagent/vminsert/vmstorage still imply possible data loss on hard stops. However, it provides stronger guarantees than auto-commit, since commits are based on TryPush rather than poll(). Fixes https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10395	2026-02-06 13:17:14 +01:00
Aliaksandr Valialkin	04f4a28cf4	docs/victoriametrics/integrations/zabbixconnector.md: add an alias - https://docs.victoriametrics.com/victoriametrics/integrations/zabbix/ - seen in the Internet Visits to this page are seen in Google Analytics reports.	2026-02-05 23:52:08 +01:00
Aliaksandr Valialkin	7f3d370244	deployment/docker: update base Alpine Docker image from 3.23.2 to 3.23.3 See https://www.alpinelinux.org/posts/Alpine-3.20.9-3.21.6-3.22.3-3.23.3-released.html	2026-02-05 19:48:54 +01:00
Aliaksandr Valialkin	c89b7f7ad5	deployment/docker: update Go builder from Go1.25.6 to Go1.25.7 See https://github.com/golang/go/issues?q=milestone%3AGo1.25.7%20label%3ACherryPickApproved	2026-02-05 19:46:56 +01:00
Aliaksandr Valialkin	d9dabea303	docs/victoriametrics: add links on how to tune VictoriaMetrics for IoT and industrial monitoring cases with low churn rate for time series The link is https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#index-tuning-for-low-churn-rate Put this link to the docs which mention IoT and industrial monitoring, so users could figure out how to optimize VictoriaMetrics for these cases.	2026-02-05 17:23:10 +01:00
Aliaksandr Valialkin	09d2ce36e8	lib/protoparser/protoparserutil: do not store byte slices with more than 75% of unused space in the pool Keeping such byte slices in the pool may increase memory usage when processing a small share of requests with much bigger sizes than the average processed request. This should help reducing memory usage at https://github.com/VictoriaMetrics/VictoriaLogs/issues/1042	2026-02-04 15:31:21 +01:00
Max Kotliar	08755c838b	docs: update changelog with LTS release notes	2026-02-02 18:45:07 +02:00
Max Kotliar	d2e438ef41	docs: bump version to v1.135.0	2026-02-02 18:38:03 +02:00
Max Kotliar	e508fa5fe2	deplyoment/docker: bump version to v1.135.0	2026-02-02 18:27:28 +02:00
f41gh7	9a7deca207	follow-up for `60cadfbad1` Respect the default value of http.DefaultTransport.Proxy. Previously, it could be unintentionally overridden with a nil value. This commit aligns Proxy configuration across all created transports.	2026-02-02 16:39:52 +01:00
Zane DeGraffenried	60cadfbad1	lib/promauth: fix oauth http client overwriting default proxy with nil Previously, default `Proxy` was unconditionally replaced with config value, which could be nil. It made impossible to use default http client proxy env variables. This commit adds check in oauth http client builder that only overwrites the transport proxy if a custom proxy url function is defined. Fixes https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10385	2026-02-02 15:39:46 +01:00
Vadim Alekseev	b36c8b1110	app/vminsert/common: reduce allocations when writing metadata Bug was introduced at `5a587f2006`, while porting change from cluster branch. This commit properlyslice `mms []metricsmetadata.Row` slice . Previously, every WriteMetadata call triggered a slice allocation. This shouldn't significantly impact overall performance, so I haven't included benchmarks. Related PR https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10392	2026-02-02 15:36:33 +01:00
Nikolay	90f0405b11	lib/promscrape: properly expose kubernetes_sd dialer metrics (#10381 ) Commit `35b31f904d` introduced a bug, where dialer metrics for Kubernetes discovery were overwritten. Fixes https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10382	2026-02-02 14:47:50 +01:00
Nikolay	eac0a7ed86	docs: mention downsampling export API behavior Fixes https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10326	2026-02-02 14:44:40 +01:00