mirror of https://github.com/VictoriaMetrics/VictoriaMetrics.git synced 2024-11-21 14:44:00 +00:00

Zakhar Bessarab ae5b85966a

lib/promscrape/discovery/dockerswarm: fix discovery filters being applied to all objects (#3632 )

* lib/promscrape/discovery/dockerswarm: fix discovery filters being applied to all objects

Signed-off-by: Zakhar Bessarab <z.bessarab@victoriametrics.com>

* Update docs/CHANGELOG.md

Signed-off-by: Zakhar Bessarab <z.bessarab@victoriametrics.com>
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>

2023-01-11 22:56:40 -08:00

326 KiB

Raw Blame History

sort
16

CHANGELOG

The following tip changes can be tested by building VictoriaMetrics components from the latest commits according to the following docs:

tip

BUGFIX: vmagent: dockerswarm_sd_configs: apply filters only to objects of the specified role. Previously filters were applied to all the objects, which could cause errors when different types of objects were used with filters that were not compatible with them. See this issue.

v1.86.1

Released at 2023-01-10

BUGFIX: return correct query results over time series with gaps. The issue has been introduced in v1.86.0.
BUGFIX: properly take into account the timeout passed by vmselect to vmstorage during query execution. This issue could result in the following error logs at vmstorage under load: cannot process vmselect request: cannot execute "search_v7": couldn't start executing the request in 0.000 seconds, since -search.maxConcurrentRequests=... concurrent requests are already executed. The issue has been introduced in v1.86.0.

v1.86.0

Released at 2023-01-10

It is recommended upgrading to VictoriaMetrics v1.86.1 because v1.86.0 contains a bug, which could lead to incorrect query results over time series with gaps.

Update note 1: This release changes the logic behind -maxConcurrentInserts command-line flag. Previously this flag was limiting the number of concurrent connections established from clients, which send data to VictoriaMetrics. Some of these connections could be temporarily idle. Such connections do not take significant CPU and memory resources, so there is no need in limiting their count. The new logic takes into account only those connections, which actively ingest new data to VictoriaMetrics and to vmagent. This means that the default -maxConcurrentInserts value should handle cases, which could require increasing the value in the previous releases. So it is recommended trying to remove the explicitly set -maxConcurrentInserts command-line flag after upgrading to this release and verifying whether this reduces CPU and memory usage.

Update note 2: The vm_concurrent_addrows_current and vm_concurrent_addrows_capacity metrics exported by vmstorage are replaced with vm_concurrent_insert_current and vm_concurrent_insert_capacity metrics in order to be consistent with the corresponding metrics exported by vminsert. Please update queries in dahsboards and alerting rules with new metric names if old metric names are used there.

FEATURE: vmagent: add support for aggregation of incoming samples by time and by labels. See these docs and this feature request.
FEATURE: vmagent: reduce memory usage when scraping big number of targets without the need to enable stream parsing mode.
FEATURE: vmagent: add support for Prometheus-compatible target discovery for HashiCorp Nomad services via nomad_sd_configs. See this feature request. Thanks to @mr-karan for the implementation.
FEATURE: vmagent: automatically pre-fetch metric_relabel_configs and the target labels when clicking on the debug metrics relabeling link at the http://vmagent:8429/targets page at the particular target. See these docs.
FEATURE: vmui: add ability to explore metrics exported by a particular job / instance. See these docs and this feature request.
FEATURE: allow passing partial RFC3339 date/time to time, start and end query args at querying APIs and export APIs. For example, 2022 is equivalent to 2022-01-01T00:00:00Z, while 2022-01-30T14 is equivalent to 2022-01-30T14:00:00Z. See these docs.
FEATURE: MetricsQL: allow using unicode letters in identifiers. For example, температура{город="Киев"} is a valid MetricsQL expression now. Previously every non-ascii letters should be escaped with \ char when used inside MetricsQL expression: \т\е\м\п\е\р\а\т\у\р\а{\г\о\р\о\д="Киев"}. Now both expressions are equivalent. Thanks to @hzwwww for the pull request.
FEATURE: relabeling: add support for keepequal and dropequal relabeling actions, which are supported by Prometheus starting from v2.41.0. These relabeling actions are almost identical to keep_if_equal and drop_if_equal relabeling actions supported by VictoriaMetrics since v1.38.0 - see these docs - so it is recommended sticking to keep_if_equal and drop_if_equal actions instead of switching to keepequal and dropequal.
FEATURE: csvimport: support empty values for imported metrics. See this issue.
FEATURE: vmalert: allow configuring the default number of stored rule's update states in memory via global -rule.updateEntriesLimit command-line flag or per-rule via rule's update_entries_limit configuration param. See these docs and this pull request.
FEATURE: improve the logic benhind -maxConcurrentInserts command-line flag. Previously this flag was limiting the number of concurrent connections from clients, which write data to VictoriaMetrics or vmagent. Some of these connections could be idle for some time. These connections do not need significant amounts of CPU and memory, so there is no sense in limiting their count. The updated logic behind -maxConcurrentInserts limits the number of active insert requests, not counting idle connections.
FEATURE: protect all the http endpoints with -httpAuth.* command-line flag. Previously endpoints protected by -*AuthKey command-line flags weren't protected by -httpAuth.*. This could complicate the proper security setup. See this issue.
FEATURE: VictoriaMetrics cluster: add -maxConcurrentInserts and -insert.maxQueueDuration command-line flags to vmstorage, so they could be tuned if needed in the same way as at vminsert nodes.
FEATURE: VictoriaMetrics cluster: limit the number of concurrently executed requests at vmstorage proportionally to the number of available CPU cores, since every request can saturate a single CPU core at vmstorage. Previously a single vmstorage could accept and start processing arbitrary number of concurrent requests received from big number of vmselect nodes. This could result in increased RAM, CPU and disk IO usage or event to out of memory crash at vmstorage side under high load. The limit can be fine-tuned if needed via -search.maxConcurrentRequests command-line flag at vmstorage according to these docs. vmstorage now exposes the following additional metrics at http://vmstorage:8482/metrics page:
- vm_vmselect_concurrent_requests_capacity - the maximum number of requests allowed to execute concurrently
- vm_vmselect_concurrent_requests_current - the current number of concurrently executed requests
- vm_vmselect_concurrent_requests_limit_reached_total - the total number of requests, which were put in the wait queue when -search.maxConcurrentRequests concurrent requests are being executed
- vm_vmselect_concurrent_requests_limit_timeout_total - the total number of canceled requests because they were sitting in the wait queue for more than -search.maxQueueDuration
BUGFIX: vmui: properly update the step value in url after the step input field has been manually changed. This allows preserving the proper step when copy-n-pasting the url to another instance of web browser. See this issue.
BUGFIX: vmui: properly update tooltip when quickly hovering multiple lines on the graph. See this issue.
BUGFIX: properly parse floating-point numbers without integer or fractional parts such as .123 and 20. during data import. See this issue.
BUGFIX: MetricsQL: properly parse durations with uppercase suffixes such as 10S, 5MS, 1W, etc. See this issue.
BUGFIX: vmagent: fix a panic during target discovery when vmagent runs with -promscrape.dropOriginalLabels command-line flag. See this issue. The bug has been introduced in v1.85.0.
BUGFIX: vmagent: dockerswarm_sd_configs: properly encode filters field. See this issue.
BUGFIX: vmagent: fix possible resource leak after hot reload of the updated consul_sd_configs. See this issue.
BUGFIX: vmagent: fix a panic in gce_sd_configs when the discovered instance has zero labels. See this issue. The issue has been introduced in v1.85.0.
BUGFIX: properly return label names starting from uppercase such as CamelCaseLabel from /api/v1/labels. See this issue.
BUGFIX: fix opentsdb HTTP endpoint not respecting -httpAuth.* flags. See this issue
BUGFIX: consistently select the sample with the biggest value out of samples with identical timestamps during querying when the deduplication is enabled according to this feature request. Previously random samples could be selected during querying.

v1.85.3

Released at 2022-12-20

Update note 1: This and newer releases of VictoriaMetrics may return gaps for rate(m[d]) queries on short time ranges if [d] lookbehind window is set expliticly. For example, rate(http_requests_total[$__interval]). This reduces confusion level when the user expects the needed results from the query with explicitly set lookbehind window. See this issue. The previous gap filling behaviour can be restored by removing explicit lookbehind window [d] from the query, e.g. by substituting the rate(m[d]) with rate(m). See these docs for details.

BUGFIX: fix error when searching for TSIDs by metricIDs in the previous indexdb: EOF error, which can occur during queries after unclean shutdown of VictoriaMetrics (e.g. via hardware reset, out of memory crash or kill -9). The error has been introduced in v1.85.2. See this issue.
BUGFIX: VictoriaMetrics enterprise: expose proper values for vm_downsampling_partitions_scheduled and vm_downsampling_partitions_scheduled_size_bytes metrics, which were added at v1.78.0. See this feature request.
BUGFIX: MetricsQL: never extend explicitly set lookbehind window for rate() function. This reduces the level of confusion when the user expects the needed results after explicitly seting the lookbehind window [d] in the query rate(m[d]). Previously VictoriaMetrics could silently extend the lookbehind window, so it covers at least two raw samples. Now this behavior works only if the lookbehind window in square brackets isn't set explicitly, e.g. in the case of rate(m). See this issue for details.
BUGFIX: vmagent: respect -usePromCompatibleNaming flag if no relabeling or extra labels were set. See this issue for details.
BUGFIX: vmui: fix the wrong legend when queries are hidden. See this issue.
BUGFIX: vmui: fix incorrect time selection after the timezone change. See this pull request.

v1.85.2

Released at 2022-12-19

FEATURE: support overriding of -search.latencyOffset value via URL param latency_offset when performing requests to /api/v1/query and /api/v1/query_range. See this issue.
FEATURE: allow changing field names in JSON logs if VictoriaMetrics components are started with -loggerFormat=json command-line flags. The field names can be changed with the -loggerJSONFields command-line flag. For example -loggerJSONFields=ts:timestamp,msg:message would rename ts and msg fields on the output JSON to timestamp and message fields. See this feature request. Thanks to @michal-kralik for the pull request.
FEATURE: vmagent: expose __meta_consul_tag_<tagname> and __meta_consul_tagpresent_<tagname> labels for targets discovered via consul_sd_configs. This simplifies converting Consul service tags to target labels with a simple relabeling rule:
```
- action: labelmap
  regex: __meta_consul_tag_(.+)
```
This resolves this StackOverflow question.
BUGFIX: properly return query results for time series, which stop receiving new samples after the rotation of indexdb. Previously such time series could be missing in query results. See this issue. The issue has been introduced in v1.83.0.
BUGFIX: allow specifying values bigger than 2GiB to the following command-line flag values on 32-bit architectures (386 and arm): -storage.minFreeDiskSpaceBytes and -remoteWrite.maxDiskUsagePerURL. Previously values bigger than 2GiB were incorrectly truncated on these architectures.
BUGFIX: vmagent: stop dropping metric name by a mistake on the /metric-relabel-debug page.

v1.85.1

Released at 2022-12-14

It is recommended upgrading to VictoriaMetrics v1.85.2 because of the bug, which may result in incomplete query results for historical time series.

FEATURE: vmalert: support $for or .For template variables in alert's annotations. See this issue.
BUGFIX: DataDog protocol parser: do not re-use host and device fields from the previously parsed messages if these fields are missing in the currently parsed message. See this issue.
BUGFIX: reduce CPU usage when the regex-based relabeling rules are applied to more than 100K unique Graphite metrics. See this issue. The issue was introduced in v1.82.0.
BUGFIX: do not block merges of small parts by merges of big parts on hosts with small number of CPU cores. This issue could result in the increasing number of storage/small parts while big merge is in progress. This, in turn, could result in increased CPU usage and memory usage during querying, since queries need to inspect bigger number of small parts. The issue has been introduced in v1.85.0.
BUGFIX: vmbackup: fix the The source request body for synchronous copy is too large and exceeds the maximum permissible limit (256MB) error when performing backups to Azure blob storage. See this issue.

v1.85.0

Released at 2022-12-11

It is recommended upgrading to VictoriaMetrics v1.85.2 because of the bug, which may result in incomplete query results for historical time series.

Update note 1: this release drops support for direct upgrade from VictoriaMetrics versions prior v1.28.0. Please upgrade to v1.84.0, wait until finished round 2 of background conversion line is emitted to log by single-node VictoriaMetrics or by vmstorage, and then upgrade to newer releases.

Update note 2: this release splits type="indexdb" metrics into type="indexdb/inmemory" and type="indexdb/file" metrics. This may break old dashboards and alerting rules, which contain label filter on {type="indexdb"}. Such label filter must be substituted with {type=~"indexdb.*"}, so it matches indexdb from the previous releases and indexdb/inmemory + indexdb/file from new releases. It is recommended upgrading to the latest available dashboards and alerting rules mentioned in these docs, since they already contain fixed label filters.

Update note 3: this release deprecates relabel_debug and metric_relabel_debug config options in scrape_configs. The -relabelDebug, -remoteWrite.relabelDebug and -remoteWrite.urlRelabelDebug command-line options are also deprecated. Use more powerful target-level relabel debugging and metric-level relabel debugging instead as documented here.

FEATURE: vmagent: provide enhanced target-level and metric-level relabel debugging. See these docs and this issue.
FEATURE: leave a sample with the biggest value for identical timestamps per each -dedup.minScrapeInterval discrete interval when the deduplication is enabled. See this issue.
FEATURE: add -inmemoryDataFlushInterval command-line flag, which can be used for controlling the frequency of in-memory data flush to disk. The data flush frequency can be reduced when VictoriaMetrics stores data to low-end flash device with limited number of write cycles (for example, on Raspberry PI). See this feature request.
FEATURE: expose additional metrics for indexdb and storage parts stored in memory and for indexdb parts stored in files (see storage docs for technical details):
- vm_active_merges{type="storage/inmemory"} - active merges for in-memory storage parts
- vm_active_merges{type="indexdb/inmemory"} - active merges for in-memory indexdb parts
- vm_active_merges{type="indexdb/file"} - active merges for file-based indexdb parts
- vm_merges_total{type="storage/inmemory"} - the total merges for in-memory storage parts
- vm_merges_total{type="indexdb/inmemory"} - the total merges for in-memory indexdb parts
- vm_merges_total{type="indexdb/file"} - the total merges for file-based indexdb parts
- vm_rows_merged_total{type="storage/inmemory"} - the total rows merged for in-memory storage parts
- vm_rows_merged_total{type="indexdb/inmemory"} - the total rows merged for in-memory indexdb parts
- vm_rows_merged_total{type="indexdb/file"} - the total rows merged for file-based indexdb parts
- vm_rows_deleted_total{type="storage/inmemory"} - the total rows deleted for in-memory storage parts
- vm_assisted_merges_total{type="storage/inmemory"} - the total number of assisted merges for in-memory storage parts
- vm_assisted_merges_total{type="indexdb/inmemory"} - the total number of assisted merges for in-memory indexdb parts
- vm_parts{type="storage/inmemory"} - the total number of in-memory storage parts
- vm_parts{type="indexdb/inmemory"} - the total number of in-memory indexdb parts
- vm_parts{type="indexdb/file"} - the total number of file-based indexdb parts
- vm_blocks{type="storage/inmemory"} - the total number of in-memory storage blocks
- vm_blocks{type="indexdb/inmemory"} - the total number of in-memory indexdb blocks
- vm_blocks{type="indexdb/file"} - the total number of file-based indexdb blocks
- vm_data_size_bytes{type="storage/inmemory"} - the total size of in-memory storage blocks
- vm_data_size_bytes{type="indexdb/inmemory"} - the total size of in-memory indexdb blocks
- vm_data_size_bytes{type="indexdb/file"} - the total size of file-based indexdb blocks
- vm_rows{type="storage/inmemory"} - the total number of in-memory storage rows
- vm_rows{type="indexdb/inmemory"} - the total number of in-memory indexdb rows
- vm_rows{type="indexdb/file"} - the total number of file-based indexdb rows
FEATURE: DataDog parser: add device tag when it is passed in the device field is present in the series object of the input request. Thanks to @PerGon for the provided pull request.
FEATURE: vmagent: improve service discovery performance when discovering big number of targets (10K and more).
FEATURE: vmagent: allow using series_limit option for limiting the number of series a single scrape target generates in stream parsing mode. See this feature request.
FEATURE: vmagent: allow using sample_limit option for limiting the number of metrics a single scrape target can expose in every response sent over stream parsing mode.
FEATURE: vmagent: add exported_ prefix to metric names exported by scrape targets if these metric names clash with automatically generated metrics such as up, scrape_samples_scraped, etc. This prevents from corruption of automatically generated metrics. See this issue.
FEATURE: vmagent: make the host label optional in DataDog data ingestion protocol. See this issue.
FEATURE: VictoriaMetrics cluster: improve error message when the requested path cannot be properly parsed, so users could identify the issue and properly fix the path. Now the error message links to url format docs. See this issue.
FEATURE: VictoriaMetrics enterprise cluster: add -storageNode.discoveryInterval command-line flag to vmselect and vminsert to control load on DNS servers when automatic discovery of vmstorage nodes is enabled. See this issue.
FEATURE: VictoriaMetrics enterprise cluster: allow reading and updating the list of vmstorage nodes at vmselect and vminsert nodes via file. See automatic discovery of vmstorage for details.
FEATURE: vmalert: reduce memory and CPU usage by up to 50% on setups with thousands of recording/alerting groups. See this issue.
FEATURE: vmalert: add -remoteWrite.sendTimeout command-line flag, which allows configuring timeout for sending data to -remoteWrite.url. See this issue.
FEATURE: vmctl: add ability to migrate data between VictoriaMetrics clusters with automatic tenants discovery. See these docs and this issue.
FEATURE: vmctl: add ability to copy data from sources via Prometheus remote_read protocol. See these docs. The related issues: one and two.
FEATURE: vmui: allow changing timezones for the requested data. See this issue.
FEATURE: vmui: provide fast path for hiding results for all the queries except the given one by clicking eye icon with ctrl key pressed. See this feature request.
FEATURE: MetricsQL: add range_trim_spikes(phi, q) function for trimming phi percent of the largest spikes per each time series returned by q. See these docs.
FEATURE: MetricsQL: allow passing inf arg into limitk, topk, bottomk and other functions, which accept numeric arg, which limits the number of output time series. See this feature request.
FEATURE: vmgateway: add support for JWT token signature verification. See these docs for details.
FEATURE: put the version of VictoriaMetrics in the first message of query trace. This should simplify debugging.
BUGFIX: vmagent: fix the The request did not have a subscription or a valid tenant level resource provider error when discovering Azure targets with azure_sd_configs. See this issue.
BUGFIX: vmalert: properly pass HTTP headers during the alert state restore procedure. See this issue.
BUGFIX: vmalert: properly specify rule evaluation step during the replay mode. The step value was previously overriden by -datasource.queryStep command-line flag.
BUGFIX: vmalert: properly return the error message from remote-write failures. Before, error was ignored and only vmalert_remotewrite_errors_total was incremented.
BUGFIX: vmui: fix sticky tooltip sizing, which could prevent from closing the tooltip. See this issue.
BUGFIX: vmui: properly put multi-line queries in the url, so it could be copy-n-pasted and opened without issues in a new browser tab. Previously the url for multi-line query couldn't be opened. See this issue.
BUGFIX: vmui: correctly handle up and down keypresses when editing multi-line queries. See this issue.

v1.84.0

Released at 2022-11-25

It is recommended upgrading to VictoriaMetrics v1.85.2 because of the bug, which may result in incomplete query results for historical time series.

FEATURE: add support for Pushgateway data import format via /api/v1/import/prometheus url. See these docs and this issue. Thanks to @PerGon for the intial implementation.
FEATURE: VictoriaMetrics cluster: add http://<vmselect>:8481/admin/tenants API endpoint for returning a list of registered tenants. See these docs for details.
FEATURE: VictoriaMetrics enterprise: add -storageNode.filter command-line flag for filtering the discovered vmstorage nodes with arbitrary regular expressions. See this feature request.
FEATURE: MetricsQL: allow using numeric values with K, Ki, M, Mi, G, Gi, T and Ti suffixes inside MetricsQL queries. For example 8Ki equals to 8*1024, while 8.2M equals to 8.2*1000*1000.
FEATURE: MetricsQL: add range_normalize function for normalizing multiple time series into [0...1] value range. This function is useful for correlation analysis of time series with distinct value ranges. See this issue.
FEATURE: MetricsQL: add range_linear_regression function for calculating simple linear regression over the input time series on the selected time range. This function is useful for predictions and capacity planning. For example, range_linear_regression(process_resident_memory_bytes) can predict future memory usage based on the past memory usage.
FEATURE: MetricsQL: add range_stddev and range_stdvar functions.
FEATURE: MetricsQL: optimize expr1 op expr2 query when expr1 returns an empty result. In this case there is no sense in executing expr2 for op not equal to or, since the end result will be empty according to PromQL series matching rules. See this issue. Thanks to @jianglinjian for pointing to this case.
FEATURE: vmui: add the ability to upload/paste JSON to investigate the trace. See this issue and this pull request.
FEATURE: vmui: reduce JS bundle size from 200Kb to 100Kb. See this pull request.
FEATURE: vmui: add the ability to hide results of a particular query by clicking the eye icon. See this pull request.
FEATURE: vmui: add copy button to row on Table view. The button copies row in MetricQL format. See this issue.
FEATURE: vmui: add compact table view. See this issue.
FEATURE: vmui: add the ability to "stick" a tooltip on the chart by clicking on a data point. See this issue and this pull request
FEATURE: vmui: add the ability to set up series custom limits. See this issue.
FEATURE: vmalert: add default alert list for vmalert's metrics. See alerts-vmalert.yml.
FEATURE: vmagent: expose vmagent_relabel_config_*, vm_relabel_config_* and vm_promscrape_config_* metrics for tracking relabel and scrape configuration hot-reloads. See this issue.
BUGFIX: MetricsQL: properly return an empty result from limit_offset if the offset arg exceeds the number of inner time series. See this issue.
BUGFIX: vmagent: properly discover GCE zones when filter option is set at gce_sd_configs. See this issue.
BUGFIX: vmui: properly display the requested graph on the requested time range when navigating from Prometheus URL in Grafana.
BUGFIX: vmui: properly display wide tables. See this issue.
BUGFIX: reduce CPU usage spikes and memory usage spikes under high data ingestion rate introduced in v1.83.0. See this issue.

v1.83.1

Released at 2022-11-10

It is recommended upgrading to VictoriaMetrics v1.85.2 because of the bug, which may result in incomplete query results for historical time series.

FEATURE: vmagent: expose __meta_consul_partition label for targets discovered via consul_sd_configs in the same way as Prometheus 2.40 does.
FEATURE: vmui: show the query trace in JSON view. See this issue. Thanks to @michal-kralik for the pull request.
BUGFIX: VictoriaMetrics enterprise: fix a panic at vminsert when the discovered list of vmstorage nodes is changed during automatic vmstorage discovery. See this issue.
BUGFIX: properly register new time series in per-day inverted index if they were ingested during the last 10 seconds of the day. See this issue. Thanks to @lmarszal for the bugreport and for the initial fix.
BUGFIX: reduce the increased memory usage spikes for some workloads. The issue was introduced in v1.83.0.
BUGFIX: properly accept OpenTSDB telnet put lines without tags without the need to specify the trailing whitespace. See this issue.

v1.83.0

Released at 2022-10-29

It is recommended upgrading to VictoriaMetrics v1.85.2 because of the bug, which may result in incomplete query results for historical time series.

Update note 1: the indexdb/tagFilters cache type at /metrics has been renamed to indexdb/tagFiltersToMetricIDs in order to make its puropose more clear.

Update note 2: vmalert: the crlfEscape template function becames obsolete starting from this release. It can be safely removed from alerting templates, since \n chars are properly escaped with other *Escape functions now. See this and this issue for details.

FEATURE: VictoriaMetrics enterprise: add support for automatic vmstorage nodes discovering and updating at vmselect and vminsert. See these docs.
FEATURE: VictoriaMetrics enterprise: allow configuring multiple retentions for distinct sets of time series. See these docs, this and this feature request.
FEATURE: VictoriaMetric cluster enterprise: add support for multiple retentions for distinct tenants - see these docs and this and this feature request.
FEATURE: allow limiting memory usage on a per-query basis with -search.maxMemoryPerQuery command-line flag. See this feature request.
FEATURE: allow referring environment variables inside command-line flags via %{ENV_VAR} syntax. For example, if AUTH_KEY=top-secret environment variable is set, then -metricsAuthKey=%{AUTH_KEY} command-line flag is automatically expanded to -storageDataPath=top-secret at VictoriaMetrics startup. See these docs for details.
FEATURE: allow referring environment variables inside other environment variables via %{ENV_VAR} syntax. For example, if A=a-%{B}, B=b-%{C} and C=c env vars are set, then VictoriaMetrics components automatically expand them to A=a-b-c, B=b-c and C=c on startup.
FEATURE: vmagent: drop all the labels with __ prefix from discovered targets in the same way as Prometheus does according to this article. Previously the following labels were available during metric-level relabeling: __address__, __scheme__, __metrics_path__, __scrape_interval__, __scrape_timeout__, __param_*. Now these labels are available only during target-level relabeling. This should reduce CPU usage and memory usage for vmagent setups, which scrape big number of targets.
FEATURE: vmagent: improve the performance for metric-level relabeling, which can be applied via metric_relabel_configs section at scrape_configs, via -remoteWrite.relabelConfig or via -remoteWrite.urlRelabelConfig command-line options.

FEATURE: vmagent: allow specifying full url in scrape target addresses (aka __address__ label). This makes valid the following -promscrape.config:

scrape_configs:
- job_name: abc
  metrics_path: /foo/bar
  scheme: https
  static_configs:
  - targets:
    # the following targets are scraped by the provided full urls
    - 'http://host1/metric/path1'
    - 'https://host2/metric/path2'
    - 'http://host3:1234/metric/path3?arg1=value1'
    # the following target is scraped by <scheme>://host4:1234<metrics_path>
    - host4:1234

See the corresponding issue.

FEATURE: vmagent: allow controlling staleness tracking on a per-scrape_config basis by specifying no_stale_markers: true or no_stale_markers: false option in the corresponding scrape_config.
FEATURE: vmalert: add strvalue and stripDomain template functions in order to improve compatibility with Prometheus.
FEATURE: vmalert: add jsonEscape and htmlEscape template functions.
FEATURE: vmui: limit the number of plotted series. This should prevent from browser crashes or hangs when the query returns big number of time series. See this feature request.
FEATURE: vmui: reduce memory usage when querying big number of time series. See this issue.
FEATURE: vmui: add responsive styles for small screens. See this issue and this pull request.
FEATURE: log error if some environment variables referred at -promscrape.config via %{ENV_VAR} aren't found. This should prevent from silent using incorrect config files.
FEATURE: immediately shut down VictoriaMetrics apps on the second SIGINT or SIGTERM signal if they couldn't be finished gracefully for some reason after receiving the first signal.
FEATURE: improve the performance of /api/v1/series endpoint by eliminating loading of unused TSID data during the API call.
FEATURE: vmbackupmanager: add functionality for automated restore from backup. See these docs.
BUGFIX: MetricsQL: properly merge buckets with identical le values, but with different string representation of these values when calculating histogram_quantile and histogram_share. For example, http_request_duration_seconds_bucket{le="5"} and http_requests_duration_seconds_bucket{le="5.0"}. Such buckets may be returned from distinct targets. Thanks to @647-coder for the pull request.
BUGFIX: vmalert: change severity level for log messages about failed attempts for sending data to remote storage from error to warn. The message for about all failed send attempts remains at error severity level.
BUGFIX: vmalert: fix panic if vmalert runs with -clusterMode command-line flag in multitenant mode. The issue has been introduced in v1.82.0.
BUGFIX: vmalert: properly escape string passed to quotesEscape template function, so it can be safely embedded into JSON string. This makes obsolete the crlfEscape function. See this and this issue.
BUGFIX: vmagent: do not show invalid error message in Kubernetes service discovery: cannot parse WatchEvent json response: EOF. The invalid error message has been appeared in v1.82.0.
BUGFIX: vmagent: properly add exported_ prefix to metric labels, which clashing with scrape target labels if honor_labels: true option isn't set in scrape_config. Previously some exported_ prefixes were missing in the resulting metric labels. See this issue. The issue has been introduced in v1.82.0.
BUGFIX: vmselect: expose missing metric vm_cache_size_max_bytes{type="promql/rollupResult"} . This metric is used for monitoring rollup cache usage with the query vm_cache_size_bytes{type="promql/rollupResult"} / vm_cache_size_max_bytes{type="promql/rollupResult"} in the same way as this is done for other cache types.

v1.82.1

Released at 2022-10-14

BUGFIX: vmui: automatically update graph, legend and url after the removal of query field. See this feature request and this comment.
BUGFIX: vmalert: remove duplicate alertname JSON entry from generated alerts. See this issue. Thanks to @Howie59 for the fix!
BUGFIX: vmalert: fix integration with Grafana via -vmalert.proxyURL, which has been broken in v1.82.0. See this issue.
BUGFIX: vmbackup: set default region to us-east-1 if AWS_REGION environment variable isn't set. The issue was introduced in vmbackup v1.82.0. See this pull request.
BUGFIX: vmbackupmanager: fix deletion of old backups at Azure blob storage.
BUGFIX: MetricsQL: properly apply regex filters when searching for time series. Previously unexpected time series could be returned from regex filter. See this issue. The issue was introduced in v1.82.0.
BUGFIX: vmagent: properly apply if section with regex filters. Previously unexpected metrics could be returned from if section. The issue was introduced in v1.82.0.

v1.82.0

Released at 2022-10-07

It isn't recommended to use VictoriaMetrics and vmagent v1.82.0 because of the bug, which may result in incorrect query results and relabeling results. Upgrade to v1.82.1 instead.

Update note 1: this release changes data format for /api/v1/export/native in incompatible way, so it cannot be imported into older version of VictoriaMetrics via /api/v1/import/native.

Update note 2: vmalert changes default value for command-line flag -datasource.queryStep from 0s to 5m. The change supposed to improve reliability of the rules evaluation when evaluation interval is lower than scraping interval.

Update note 3: vm_account_id and vm_project_id labels must be passed to tcp-based Graphite, InfluxDB and OpenTSDB endpoints at VictoriaMetrics cluster instead of undocumented VictoriaMetrics_AccountID and VictoriaMetrics_ProjectID labels when writing samples to the needed tenant. See these docs for details.

FEATURE: VictoriaMetrics cluster: support specifying tenant ids via vm_account_id and vm_project_id labels. See these docs and this feature request.
FEATURE: vmagent: improve relabeling performance by up to 3x for non-trivial regex values such as ([^:]+):.+, which can be used for extracting a host part from host:port label value.
FEATURE: MetricsQL: improve performance by up to 4x for queries containing non-trivial regex filters such as {path=~"/foo/.+|/bar"}.
FEATURE: improve performance scalability on systems with many CPU cores for /federate and /api/v1/export/... endpoints.
FEATURE: sanitize metric names for data ingested via DataDog protocol according to DataDog metric naming. The behaviour can be disabled by passing -datadog.sanitizeMetricName=false command-line flag. Thanks to @PerGon for the pull request.
FEATURE: add -usePromCompatibleNaming command-line flag to vmagent, to single-node VictoriaMetrics and to vminsert component of VictoriaMetrics cluster. This flag can be used for normalizing the ingested metric names and label names to Prometheus-compatible form. If this flag is set, then all the chars unsupported by Prometheus are replaced with _ chars in metric names and labels of the ingested samples. See this feature request.
FEATURE: accept whitespace in metric names and tags ingested via Graphite plaintext protocol according to the specs. See this issue.
FEATURE: check the correctess of raw sample timestamps stored on disk when reading them. This reduces the probability of possible silent corruption of the data stored on disk. This should help this and this issue.
FEATURE: atomically delete directories with snapshots, parts and partitions at storage level. Previously such directories can be left in partially deleted state when the deletion operation was interrupted by unclean shutdown. This may result in cannot open file ...: no such file or directory error on the next start. The probability of this error was quite high when NFS or EFS was used as persistent storage for VictoriaMetrics data. See this issue.
FEATURE: set the start arg to end - 5 minutes if isn't passed explicitly to /api/v1/labels and /api/v1/label/.../values. See this pull request.
FEATURE: allow to define the minimum TLS version to use when accepting https requests to VictoriaMetrics components if -tls command-line flag is set. The minimum TLS version can be set via -tlsMinVersion command-line flag. See this feature request.
FEATURE: vmctl: add vm-native-step-interval command line flag for vm-native mode. New option allows splitting the import process into chunks by time interval. This helps migrating data sets with high churn rate and provides better control over the process. See feature request.
FEATURE: vmui: add top queries tab, which shows various stats for recently executed queries. See these docs and this feature request.
FEATURE: vmui: move the "Execute Query" and "Add Query" buttons below the query fields, change icon for remove query. See this issue.
FEATURE: vmui: set the maximum number of queries to 4, remove multi Y-axes, left one for all queries and dotted lines to indicate queries in the graph. See this issue.
FEATURE: vmalert: add debug mode to the alerting rule settings for printing additional information into logs during evaluation. See debug param in alerting rule config.
FEATURE: vmalert: add experimental feature for displaying last 10 states of the rule (recording or alerting) evaluation. The state is available on the Rule page, which can be opened by clicking on Details link next to Rule's name on the /groups page.
FEATURE: vmalert: allow using extra labels in annotiations. See this feature request.
FEATURE: vmalert: allow configuring authorization params per list of targets in vmalert's notifier config for static_configs. See this issue.
FEATURE: vmalert: allow using {% raw %}{{$labels}}{% endraw %} for templating in command-line flag -external.alert.source. The change supposed to provide additional flexibility for generating alert's source link based on labels values.
FEATURE: vmalert: add vm_account_id and vm_project_id labels to results of alerting and recording rules if -clusterMode is enabled. This improves multitenant support in vmalert.
FEATURE: vmagent: minimize the time needed for reading large responses from scrape targets in stream parsing mode. This should reduce scrape durations for such targets as kube-state-metrics running in a big Kubernetes cluster.
FEATURE: MetricsQL: add sort_by_label_numeric and sort_by_label_numeric_desc functions for numeric sort of input time series by the specified labels. See this feature request.
FEATURE: vmbackup and vmrestore: retry GCS operations for up to 3 minutes on temporary failures. See this issue.
FEATURE: vmbackup: add support for saving / restoring backups to / from Azure blob storage. See this feature request.
FEATURE: vmbackupmanager: expose vm_backup_in_flight metric, which can be used for determining which backup types - latest, hourly, daily, weekly or monthly - are currently executed.
FEATURE: vmgateway: add ability to extract JWT authorization token from non-standard HTTP header by passing it via -auth.httpHeader command-line flag. See this feature request.
FEATURE: vmagent: expose __meta_ec2_region label for ec2_sd_config in the same way as Prometheus 2.39 does.
FEATURE: vmagent: accept data ingestion requests via paths starting from /prometheus prefix in the same way as VictoriaMetrics does. For example, vmagent now accepts Prometheus remote_write data via both /api/v1/write and /prometheus/api/v1/write. This simplifies switching between single-node VictoriaMetrics and vmagent.
FEATURE: vmagent: add external_labels from global section at -promscrape.config after the relabeling is applied to scraped metrics. This aligns with Prometheus behaviour. Previously the external_labels were added to scrape targets, so they could be modified during relabeling. See this issue.
FEATURE: vmagent: allow specifying per--remoteWrite.url limits for on-disk size for pending data via -remoteWrite.maxDiskUsagePerURL command-line flag. Thanks to @rbizos for the pull request.
FEATURE: VictoriaMetrics cluster: log clear error when multiple identical -storageNode command-line flags are passed to vmselect or to vminsert. Previously these components were crashed with cryptic panic metric ... is already registered in this case. See this issue.
BUGFIX: do not export stale metrics via /federate api after the staleness markers. Previously such metrics were exported with NaN values. this could break some setups. See this issue.
BUGFIX: export ininity numbers as "Infinity" strings at /api/v1/export, so they can be parsed by standard JSON parsers. Previously infinity numbers were exported as Inf values, which couldn't be parsed by standard JSON parsers. See this issue.
BUGFIX: vmauth: properly handle request paths ending with / such as /vmui/. Previously vmui was dropping the traling /, which could prevent from using vmui via vmauth. See this issue.
BUGFIX: vmagent: properly encode query params for aws signed requests, use %20 instead of + as api requires. See this issue.
BUGFIX: vmagent: properly parse relabel config when regex ending with escaped $. See this issue.
BUGFIX: MetricsQL: properly calculate rate_over_sum(m[d]) as sum_over_time(m[d])/d. Previously the sum_over_time(m[d]) could be improperly divided by smaller than d time range. See rate_over_sum() docs and this issue.
BUGFIX: MetricsQL: properly calculate increase(m[d]) over slow-changing counters with values smaller than 100. Previously increase could return unexpectedly big results in this case. See the related issue and this pull request.
BUGFIX: MetricsQL: ignore empty series when applying limit_offset. It should improve queries with additional filters by value in expressions like limit_offset(1,1, foo > 1).
BUGFIX: MetricsQL: properly calculate quantiles_over_time when the lookbehind window contains only a single sample. Previously an empty result was incorrectly returned in this case.
BUGFIX: vmui: fix RangeError: Maximum call stack size exceeded error when the query returns too many data points at Table view. See this pull request.
BUGFIX: vmui: fix workaround for adding more queries via URL. See this issue.
BUGFIX: vmalert: re-evaluate annotations per each alert evaluation. Previously, annotations were evaluated only on alert's value change. This could result in stale annotations in some cases described in this pull request.
BUGFIX: prevent from excessive CPU usage when the storage enters read-only mode. The previous fix in v1.81.0 wasn't complete.
BUGFIX: vmalert: change default value for command-line flag -datasource.queryStep from 0s to 5m. Param step is added by vmalert to every rule evaluation request sent to datasource. Before this change, step was equal to group's evaluation interval by default. Param step for instant queries defines how far VM can look back for the last written data point. The change supposed to improve reliability of the rules evaluation when evaluation interval is lower than scraping interval.
BUGFIX: properly calculate vm_rows_scanned_per_query histogram exported at /metrics page of vmselect and single-node VictoriaMetrics. Previously it could return misleadingly high numbers for rollup functions, which scan only a few samples on the provided lookbehind window in square brackets. For example, increase(m[1d]) always scans only 2 rows (aka raw samples) per each returned time series.

v1.81.2

Released at 2022-09-08

BUGFIX: VictoriaMetrics cluster: properly calculate query results at vmselect. See this issue. The issue has been introduced in v1.81.0.

v1.81.1

Released at 2022-09-02

It isn't recommended to use VictoriaMetrics cluster v1.81.1 because of the bug, which may result in incorrect query results. Upgrade to v1.81.2 instead.

FEATURE: MetricsQL: evaluate q1, ..., qN in parallel when calculating union(q1, .., qN). Previously union args were evaluated sequentially. This could result in lower than expected performance.
BUGFIX: VictoriaMetrics cluster: fix potential panic at vmselect under high load, which has been introduced in v1.81.0. See this issue.

v1.81.0

It isn't recommended to use VictoriaMetrics cluster v1.81.0 because of the bug, which may result in vmselect crashes under high load. Upgrade to v1.81.2 instead.

Released at 2022-08-31

Update note 1: vmalert by default hides values of -remoteWrite.url, -remoteRead.url and -datasource.url in logs and at http://vmalert:8880/flags for security reasons. See the corresponding SECURITY change in the Chagelog below for additional info.

Update note 2: vmalert by default points alert source url to /vmalert/alert?... aka web UI instead of /vmalert/api/v1/alert?... aka JSON handler. The old behavior can be achieved by setting {% raw %}-external.alert.source=vmalert/api/v1/alert?group_id={{.GroupID}}&alert_id={{.AlertID}}{% endraw %} command-line flag.

SECURITY: vmalert: do not expose -remoteWrite.url, -remoteRead.url and -datasource.url command-line flag values in logs and at http://vmalert:8880/flags page by default, since they may contain sensitive data such as auth keys. This aligns vmalert behaviour with vmagent, which doesn't expose -remoteWrite.url command-line flag value in logs and at http://vmagent:8429/flags page by default. Specify -remoteWrite.showURL, -remoteRead.showURL and -datasource.showURL command-line flags for showing values for the corresponding -*.url flags in logs. Thanks to @mble for the pull request.
SECURITY: upgrade base docker image (alpine) from 3.16.1 to 3.16.2. See alpine 3.16.2 release notes.
FEATURE: return shorter error messages to Grafana and to other clients requesting /api/v1/query and /api/v1/query_range endpoints. This should simplify reading these errors by humans. The long error message with full context is still written to logs.
FEATURE: add the ability to fine-tune the number of points, which can be generated per each matching time series during subquery evaluation. This can be done with the -search.maxPointsSubqueryPerTimeseries command-line flag. See this feature request.
FEATURE: vmagent: improve the performance for relabeling rules with commonly used regular expressions in regex and if fields such as some_string, prefix.*, prefix.+, foo|bar|baz, .*foo.* and .+foo.+.
FEATURE: vmagent: reduce CPU usage when discovering big number of Kubernetes targets with big number of labels and annotations.
FEATURE: vmagent: add ability to accept multitenant data via OpenTSDB /api/put protocol at /insert/<tenantID>/opentsdb/api/put http endpoint if multitenant support is enabled at vmagent. Thanks to @chengjianyun for the pull request.
FEATURE: monitoring: expose vm_hourly_series_limit_max_series, vm_hourly_series_limit_current_series, vm_daily_series_limit_max_series and vm_daily_series_limit_current_series metrics when -search.maxHourlySeries or -search.maxDailySeries limits are set. This allows alerting when the number of unique series reaches the configured limits. See these docs for details.
FEATURE: VictoriaMetrics cluster: reduce the amounts of logging at vmstorage when vmselect connects/disconnects to vmstorage.
FEATURE: VictoriaMetrics cluster: improve performance for heavy queries on systems with many CPU cores.
FEATURE: vmagent: add ability to use {% raw %}{{label_name}}{% endraw %} placeholders in the replacement option of relabeling rules. This simplifies constructing label values from multiple existing label values. See these docs for details.
FEATURE: vmagent: generate additional per-target metrics - scrape_series_limit, scrape_series_current and scrape_series_limit_samples_dropped if series limit is set according to these docs. This simplifies alerting on targets with the exceeded series limit. See these docs for details on these metrics.
FEATURE: vmagent: add support for MX record types in dns_sd_configs in the same way as Prometheus 2.38 does.
FEATURE: vmagent: add __meta_kubernetes_service_port_number meta-label for role: service in kubernetes_sd_configs in the same way as Prometheus 2.38 does.
FEATURE: vmagent: add __meta_kubernetes_pod_container_image meta-label for role: pod in kubernetes_sd_configs in the same way as Prometheus 2.38 does.
FEATURE: vmagent: retry HTTP requests after some wait time during service discovery and during target scrapes if the server returns 429 HTTP status code (aka Too many requests). See this issue.
FEATURE: vmui: add a legend in the top right corner for shortcut keys. See this feature request.
FEATURE: vmalert: add toTime() template function in the same way as Prometheus 2.38 does. See these docs.
FEATURE: vmalert: add $alertID and $groupID template variables. These variables may be used for templating annotations or -external.alert.source command-line flag. See the full list of supported variables here.
FEATURE: vmalert: add $activeAt template variable. See this feature request. See the full list of supported variables here. Thanks to @laixintao for the pull request.
FEATURE: vmalert: point alert source to vmalert's UI at /vmalert/alert?... instead of JSON handler at /vmalert/api/v1/alert?.... This improves user experience. The old behavior can be achieved by setting {% raw %}-external.alert.source=vmalert/api/v1/alert?group_id={{.GroupID}}&alert_id={{.AlertID}}{% endraw %} command-line flag.
BUGFIX: prevent from excess CPU usage when the storage enters read-only mode.
BUGFIX: improve performance for requests to /api/v1/labels and /api/v1/label/.../values when the filter in the match[] query arg matches small number of time series. The performance for this case has been reduced in v1.78.0. See this and this issues.
BUGFIX: increase the default limit on the number of concurrent merges for small parts from 8 to 16. This should help resolving potential issues with heavy data ingestion. See this comment from @lukepalmer .
BUGFIX: MetricsQL: fix panic when incorrect arg is passed as phi into histogram_quantiles function. See this issue.

v1.80.0

Released at 2022-08-08

FEATURE: vmalert: allow configuring additional HTTP request headers for -datasource.url, -remoteWrite.url and -remoteRead.url via -datasource.headers, -remoteWrite.headers and -remoteRead.headers command-line flags. Additional HTTP request headers also can be set on group level via headers param - see these docs and this issue.
FEATURE: MetricsQL: execute left and right sides of certain operations in parallel. For example, q1 or q2, aggr_func(q1) <op> q2, q1 <op> aggr_func(q1). This may improve query performance if VictoriaMetrics has enough free resources for parallel processing of both sides of the operation. See this feature request.
FEATURE: vmauth: allow multiple sections with duplicate username but with different password values at -auth.config file.
FEATURE: add ability to push internal metrics (e.g. metrics exposed at /metrics page) to the configured remote storage from all the VictoriaMetrics components. See these docs.
FEATURE: improve performance for heavy queries over big number of time series on systems with big number of CPU cores. See this issue. Thanks to @zqyzyq for the idea.
FEATURE: improve performance for registering new time series in indexdb by up to 50%. Thanks to @ahfuzhang for the issue.
FEATURE: vmagent: add ability to specify tenantID in target labels. In this case metrics from the given target are routed to the given __tenant_id__. See these docs and this feature request.
FEATURE: vmagent: add service discovery for Yandex Cloud. See these docs and this feature request.
FEATURE: vmui. Zoom in the graph by selecting the needed time range in the same way Grafana does. Hold ctrl (or cmd on MacOS) in order to move the graph to the left/right. Hold ctrl (or cmd on MacOS) and scroll up/down in order to zoom in/out the area under the cursor. See this feature request.
BUGFIX: VictoriaMetrics cluster: fix potential panic in multi-level cluster setup when top-level vmselect is configured with -replicationFactor bigger than 1. See this issue.
BUGFIX: vmagent: properly handle custom endpoint value in ec2_sd_configs. It was ignored since v1.77.0 because of a bug in the implementation of this feature request. See this issue.
BUGFIX: vmagent: add missing __meta_kubernetes_ingress_class_name meta-label for role: ingress service discovery in Kubernetes. See this commit from Prometheus.
BUGFIX: vmagent: allow stale responses from Consul service discovery (aka consul_sd_configs) by default in the same way as Prometheus does. This should reduce load on Consul when discovering big number of targets. Stale responses can be disabled by specifying allow_stale: false option in consul_sd_config. See this issue.
BUGFIX: vmagent: dockerswarm_sd_configs: properly set __meta_dockerswarm_container_label_* labels instead of __meta_dockerswarm_task_label_* labels as Prometheus does. See this issue.
BUGFIX: vmagent: set up metric to 0 for partial scrapes in stream parsing mode. Previously the up metric was set to 1 when at least a single metric has been scraped before the error. This aligns the behaviour of vmselect with Prometheus.
BUGFIX: vmagent: restart all the scrape jobs during config reload after global section is changed inside -promscrape.config. See this issue.
BUGFIX: vmagent: properly assume role with AWS ECS credentials. See this issue. Thanks to @transacid for the fix.
BUGFIX: vmagent: do not split regex in relabeling rules into multiple lines if it contains groups. This fixes this issue.
BUGFIX: MetricsQL: return series from q1 if q2 doesn't return matching time series in the query q1 ifnot q2. Previously series from q1 weren't returned in this case.
BUGFIX: vmui: properly show date picker at Table tab. See this issue.
BUGFIX: properly generate http redirects if -http.pathPrefix command-line flag is set. See this issue.