app/vmalert: detect alerting rules which don't match any series at all
vmalert starts to understand /query responses which contain object:
```
"stats":{"seriesFetched": "42"}
```
If object is present, vmalert parses it and populates a new field
`SeriesFetched`. This field is then used to populate the new metric
`vmalert_alerting_rules_last_evaluation_series_fetched` and to
display warnings in the vmalert's UI.
If response doesn't contain the new object (Prometheus or
VictoriaMetrics earlier than v1.90), then `SeriesFetched=nil`.
In this case, UI will contain no additional warnings.
And `vmalert_alerting_rules_last_evaluation_series_fetched` will
be set to `-1`. Negative value of the metric will help to compile
correct alerting rule in follow-up.
Thanks for the initial implementation to @Haleygo
See https://github.com/VictoriaMetrics/VictoriaMetrics/pull/4056
See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/4039
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: use group's ID in UI to avoid collisions
Identical group names are allowed. So we should used IDs
for various groupings and aggregations in UI.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: prevent disabling state updates tracking
The minimum number of update states to track is now set to 1.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: properly update `debug` and `update_entries_limit` params on hot-reload
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: display `debug` field for rule in UI
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: exclude `updates` field from json marhsaling
This field isn't correctly marshaled right now.
And implementing the correct marshaling for it doesn't
seem right, since json representation is mostly used
by systems like Grafana. And Grafana doesn't expect this
field to be present.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* fix test for disabled state
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* fix test for disabled state
Signed-off-by: hagen1778 <roman@victoriametrics.com>
---------
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: speed up state restore procedure on start
Alerts state restore procedure has been changed to become asynchronous.
It doesn't block groups start anymore which significantly improves vmalert's startup time.
Instead, state restore is called by each group in their goroutines after the first rules
evaluation.
While previously state restore attempt was made for all loaded alerting rules,
now it is called only for alerts which became active after the first evaluation.
This reduces the amount of API calls to the configured remote read URL.
This also means that `remoteRead.ignoreRestoreErrors` command-line flag becomes deprecated now
and will have no effect if configured.
See relevant issue https://github.com/VictoriaMetrics/VictoriaMetrics/issues/2608
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* make lint happy
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* Apply suggestions from code review
---------
Signed-off-by: hagen1778 <roman@victoriametrics.com>
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>
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.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
http.Request was used as a part of state struct
for generating the curl command when viewing the rule's
state changes.
It appears, that holding a referencing is far more expensive
than generating the curl command immediately.
On the test with 40k rules, this change reduces memory
and CPU usage by 50%.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
Signed-off-by: hagen1778 <roman@victoriametrics.com>
According to Ruler specification, only labels returned within time series
should be available for use in annotations.
For long time, vmalert didn't respect this rule. And in PR
https://github.com/VictoriaMetrics/VictoriaMetrics/pull/2403
this was fixed for the sake of compatibility. However, this resulted
into users confusion, as they expected all configured and extra labels
to be available - https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3013
This fix allows to use extra labels in Annotations. But in the case of conflicts
the original labels (extracted from time series) are preferred.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: always re-evaluate Annotations
Previously, Annotations were evaluated only:
1. On alert creating.
2. On alert's value change.
This is premature optimization. It was assumed that since annotations
could contain only text with alert's labels or value - there is no need
in spending resources to re-compile Annotations.
Later, template function `query` was added, which can execute
arbitrary queries and return different results on every evaluation.
So if it was used in annotations, it would be executed only on init
or value change.
Another case when optimization caused an issue - annotations hot reload.
In this case, annotations of the active alert won't change even if Rule's
annotations were changed.
This fix enables Annotations re-evaluation on each iteration to resolve
issues above. It would have some impact on performance, but it is unlikely
it will be noticeable.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: add tp Changelog
Signed-off-by: hagen1778 <roman@victoriametrics.com>
Signed-off-by: hagen1778 <roman@victoriametrics.com>
The change adds an example of `curl` command to the Rule's page.
The command is generated for each recorded state. It is supposed
user can just copy&execute the command to see what was returned
to vmalert.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
vmalert: add experimental feature of storing Rule's evaluation state
The new feature keeps last 20 state changes of each Rule
in memory. The state are available for view on the Rule's
view page. The page can be opened by clicking on `Details`
link next to Rule's name on the `/groups` page.
States change suppose to help in investigating cases when Rule
doesn't generate alerts or records.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
- Rename logDebug() to logDebugf() and pass format string together
with format args directly to logDebugf(). This eliminates fmt.Sprintf()
overhead at logDebug() call site when debugging is disabled.
- Format labels in debug message in Prometheus format, e.g. {label1="value1",...labelN="valueN"}
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3025
* vmalert: add `debug` mode for alerting rules
Debug information includes alerts state changes and requests
sent to the datasource. Debug can be enabled only on rule's
level. It might be useful for debugging unexpected
behaviour of alerting rule.
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3025
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: review fixes
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* Update app/vmalert/alerting.go
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>
* vmalert: go fmt
Signed-off-by: hagen1778 <roman@victoriametrics.com>
Signed-off-by: hagen1778 <roman@victoriametrics.com>
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>
vmalert can be successfully used with datasources
compatible with Prometheus HTTP API. So we remove comments or
notes in Readme which are saying opposite.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
vmalert: support `limit` param in groups definition
`limit` param limits number of time series samples produced by a single rule
during execution.
On reaching the limit rule will return an err.
Signed-off-by: lihaowei <haoweili35@gmail.com>
* vmalert: calculate time for firing alert based on the given timestamp
Previously, current time was used for checking the `firing` threshold.
This is not correct, since alerts are evaluated at specific timestamps.
Hence, this specific timestamp supposed to be used in the calculation.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: properly calculate evaluation timestamp for rules
Timestamp for rules evaluation should be calculated after
the artifical delay for groups start. Otherwise, evaluation
timestamp can fall back too far in time.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
To improve compatibility with Prometheus alerting the order of
templates processing has changed.
Before, vmalert did all labels processing beforehand. It meant
all extra labels (such as `alertname`, `alertgroup` or rule labels)
were available in templating. All collisions were resolved in favour
of extra labels.
In Prometheus, only labels from the received metric are available in
templating, so no collisions are possible.
This change makes vmalert's behaviour similar to Prometheus.
For example, consider alerting rule which is triggered by time series
with `alertname` label. In vmalert, this label would be overriden
by alerting rule's name everywhere: for alert labels, for annotations, etc.
In Prometheus, it would be overriden for alert's labels only, but in annotations
the original label value would be available.
See more details here https://github.com/prometheus/compliance/issues/80
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: split alert's `Start` field into `ActiveAt` and `Start`
The `ActiveAt` field identifies when alert becomes active for rules
with `for > 0`. Previously, this value was stored in field `Start`.
The field `Start` now identifies the moment alert became `FIRING`.
The split is needed in order to distinguish these two moments
in the API responses for alerts.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: support specific moment of time for rules evaluation
The Querier interface was extended to accept a new argument
used as a timestamp at which evaluation should be made.
It is needed to align rules execution time within the group.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: mark disappeared series as stale
Series generated by alerting rules, which were sent to remote write
now will be marked as stale if they will disappear on the next
evaluation. This would make ALERTS and ALERTS_FOR_TIME series
more precise.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* wip
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: evaluate rules at fixed timestamp
Before, time at which rules were evaluated was calculated
right before rule execution. The change makes sure
that timestamp is calculated only once per evalution round
and all rules are using the same timestamp.
It also updates the logic of resending of already resolved
alert notification.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: allow overridin `alertname` label value if it is present in response
Previously, `alertname` was always equal to the Alerting Rule name. Now,
its value can be overriden if series in response containt the different value
for this label.
The change is needed for improving compatibility with Prometheus.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: align rules evaluation in time
Now, evaluation timestamp for rules evaluates as if
there was no delay in rules evaluation. It means, that
rules will be evaluated at fixed timestamps+group_interval.
This way provides more consistent evaluation results and
improves compatibility with Prometheus,
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: add metric for missed iterations
New metric `vmalert_iteration_missed_total` will show
whether rules evaluation round was missed.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: reduce delay before the initial rule evaluation in group
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: rollback alertname override
According to the spec:
```
The alert name from the alerting rule (HighRequestLatency from the example above) MUST be added to the labels of the alert with the label name as alertname. It MUST override any existing alertname label.
```
https://github.com/prometheus/compliance/blob/main/alert_generator/specification.md#step-3
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: throw err immediately on dedup detection
```
The execution of an alerting rule MUST error out immediately and MUST NOT send any alerts
or add samples to samples receiver if there is more than one alert with the same labels
```
https://github.com/prometheus/compliance/blob/main/alert_generator/specification.md#step-4
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: cleanup
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: use strings builder to reduce allocs
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: add support of `resendDelay` flag for alerts
Co-authored-by: dmitryk-dk <dmitry.kozlov@brightlocal.com>
Co-authored-by: hagen1778 <roman@victoriametrics.com>
* vmalert: add support for `sortByLabel` template function
* vmalert: update API according to Prometheus conformance program
The changes to the API, field names and URL path has been made
according to the Prometheus specification for `alert_generator`
https://github.com/prometheus/compliance/blob/main/alert_generator/specification.md
* vmalert: fix the timestamp of the evaluated rules
The timestamp used for alert's `EndsAt` was calculated
before sending the notification. While the correct way
is to use the timestamp taken right before rules evaluation.
* vmalert: add `-datasource.queryTimeAlignment` flag
The flag is supposed to provide ability to disable `time`
param alignment when executing rules. By default, this flag
is enabled, so it remains backward compatible.
The flag was introduced to achieve better compatibility
with Prometheus behaviour according to https://github.com/prometheus/compliance/blob/main/alert_generator/specification.md
Signed-off-by: hagen1778 <roman@victoriametrics.com>
vmalert: support configuration file for notifiers
* vmalert notifiers now can be configured via file
see https://docs.victoriametrics.com/vmalert.html#notifier-configuration-file
* add support of Consul service discovery for notifiers config
see https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1947
* add UI section for currently loaded/discovered notifiers
* deprecate `-rule.configCheckInterval` in favour of `-configCheckInterval`
* add ability to suppress logs for duplicated targets for notifiers discovery
* change behaviour of `vmalert_alerts_send_errors_total` - it now accounts
for failed alerts, not HTTP calls.
Service labels like `alertname` or `alertgroup` were attached
after template expanding for `labels` section. Because of this,
labels `alertname` or `alertgroup` weren't available for templating
in `labels` section of alert's definition.
This commit changes the order of labels attaching and adds a test
for verifying these labels availability.
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1921
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: introduce additional HTTP URL params per-group configuration
The new group field `params` allows to configure custom HTTP URL params
per each group. These params will be applied to every request before
executing rule's expression. Hot config reload is also supported.
Field `extra_filter_labels` was deprecated in favour of `params` field.
vmalert will print deprecation log message if config file contains
the deprecated field.
`params` fields are supported by both Prometheus and Graphite datasource types.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: provide more examples for `params` field
Signed-off-by: hagen1778 <roman@victoriametrics.com>
* vmalert: set higher priority for `params` setting
If there would be a conflict between URL params set in `datasource.url` flag
and params in group definition the latter will have higher priority.
Signed-off-by: hagen1778 <roman@victoriametrics.com>
Previously, ID for alert entity was generated without alertname or groupname.
This led to collision, when multiple alerting rules within the same group
producing same labelsets. E.g. expr: `sum(metric1) by (job) > 0` and
expr: `sum(metric2) by (job) > 0` could result into same labelset `job: "job"`.
The issue affects only UI and Web API parts of vmalert, because alert ID is used
only for displaying and finding active alerts. It does not affect state restore
procedure, since this label was added right before pushing to remote storage.
The change now adds all extra labels right after receiving response from the datasource.
And removes adding extra labels before pushing to remote storage.
Additionally, change introduces a new flag `Restored` which will be displayed in UI
for alerts which have been restored from remote storage on restart.
The source link is controlled by `external.url` and `external.alert.source`
flags, in the same way as for alertmanager notifications.
The source link is added to Alerts list view, and specific Alert view.
New UI pages:
/ - welcome page with API handlers list;
/groups - list of all rules per group;
/alerts - list of all active alerts;
/groupID/alertID/status - status of the active alert;
* vmalert: expose new metrics for tracking number of produced samples during last evaluation
Two new metrics were added to track the number of samples produced during the last evaluation:
* vmalert_recording_rules_last_evaluation_samples
* vmalert_alerting_rules_last_evaluation_samples
The gauge type is used to remain consistent with Prometheus metric
`prometheus_rule_group_last_evaluation_samples` which is on the group level.
However, the counter type was considered as well.
Two metrics instead of one are used to make it easier to separate recording and
alerting rules. It is likely, number of samples produced by recording rules is
more important so people will refer to it more frequently.
The expected usage of the new metric is the following:
```
- alert: RecordingRuleReturnsEmptyResults
expr: sum(vmalert_recording_rules_last_evaluation_samples) by(recording) < 1
annotations:
summary: Recording rule {{$labels.recording}} returns empty results.
Please verify expression correctness.
```
Addresses https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1494
* vmalert: rename `vmalert_alerts_error` to `vmalert_alerting_rules_error` to remain consistent with recording rules metrics
* vmalert: support rules backfilling (aka `replay`)
vmalert can `replay` configured rules in the past
and backfill results via remote write protocol.
It supports MetricsQL/PromQL storage as data source,
and can backfill data to remote write compatible
storage.
Supports recording and alerting rules `replay`. See more
details in README.
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/836
* vmalert: review fixes
* vmalert: readme fixes
The new setting `extra_filter_labels` may be assigned to group.
If it is, then all rules within a group will automatically filter
for configured labels. The feature is well-described here
https://docs.victoriametrics.com#prometheus-querying-api-enhancements
New setting is compatible only with VM datasource.
Previously, `startGroup` could exit on restore errors despite the
`remoteRead.ignoreRestoreErrors` flag value. Now vmalert checks the
flag value before deciding whether to return error or just log it.
Alerting rules now can return specific error type ErrStateRestore to indicate
whether restore state procedure failed. Such errors were returned and logged
before as well. But now user can specify whether to just log these errors
(remoteRead.ignoreRestoreErrors=true) or to stop the process
(remoteRead.ignoreRestoreErrors=false). The latter is important when VM isn't
ready yet to serve queries from vmalert and it needs to wait.
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1252
* Simplify arguments list for fn `queryDataSource` to improve readbility
* vmalert: adjust `time` param according to rule evaluation interval
With this change, vmalert will start to use rule's evaluation interval
for truncating the `time` param. This is mostly needed to produce consistent
time series with timestamps unaffected by vmalert start time. Now, timestamp
becomes predictable.
Additionally, adjustment is similar to what Grafana does for plotting range graphs.
Hence, recording rule series and recording rule expression plotted in grafana
suppose to become similar in most of cases.
* changes vmalert Querier with per rule querier
it allows to changes some parametrs based on rule setting
for instance - alert type, tenant for cluster version or event endpoint url.
* init implementation for graphite alerts
* adds graphite support for vmalert
* small fix
* changes vmalert graphite api with type
* updates tests
* small fix
* fixes graphite parse
* Fixes graphite from time
The commit adds a support for template function `query`,
`first` and `value`. The function `query` executes
a MetricsQL query for active alerts. In vmalert we
update templates on every evaluation for active alerts
to keep them up to date. With `query` func it may become
a perf issue since it will fire a query on every execution.
We should keep it in mind for now.
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/539
The previous implementation treated extra labels (global and rule labels) as
separate label set to returned time series labels. Hence, time series always contained
only original labels and alert ID was generated from sorted labels key-values.
Extra labels didn't affect the generated ID and were applied on the following actions:
- templating for Summary and Annotations;
- persisting state via remote write;
- restoring state via remote read.
Such behaviour caused difficulties on restore procedure because extra labels had to be dropped
before checking the alert ID, but that not always worked. Consider the case when expression
returns the following time series `up{job="foo"}` and rule has extra label `job=bar`.
This would mean that restored alert ID will be always different to the real time series because
of collision.
To solve the situation extra labels are now always applied beforehand and `vmalert` doesn't
store original labels anymore. However, this could result into a new error situation.
Consider the case when expression returns two time series `up{job="foo"}` and `up{job="baz"}`,
while rule has extra label `job=bar`. In such case, applying extra labels will result into
two identical time series and `vmalert` will return error:
`result contains metrics with the same labelset after applying rule labels`
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/870
Label `alertgroup` was introduced in #611 and automatically added to generated
time series. By mistake, this new label wasn't correctly purged on restore event
and affected alert's ID uniqueness. This commit removes `alertgroup` label
in restore function.
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/870
* app/vmalert: extend metrics set exported by `vmalert` #573
New metrics were added to improve observability:
+ vmalert_alerts_pending{alertname, group} - number of pending alerts per group
per alert;
+ vmalert_alerts_acitve{alertname, group} - number of active alerts per group
per alert;
+ vmalert_alerts_error{alertname, group} - is 1 if alertname ended up with error
during prev execution, is 0 if no errors happened;
+ vmalert_recording_rules_error{recording, group} - is 1 if recording rule
ended up with error during prev execution, is 0 if no errors happened;
* vmalert_iteration_total{group, file} - now contains group and file name labels.
This should improve control over specific groups;
* vmalert_iteration_duration_seconds{group, file} - now contains group and file name labels. This should improve control over specific groups;
Some collisions for alerts and recording rules are possible, because neither
group name nor alert/recording rule name are unique for compatibility reasons.
Commit contains list of TODOs for Unregistering metrics since groups and rules
are ephemeral and could be removed without application restart. In order to
unlock Unregistering feature corresponding PR was filed - https://github.com/VictoriaMetrics/metrics/pull/13
* app/vmalert: extend metrics set exported by `vmalert` #573
The changes are following:
* add an ID label to rules metrics, since `name` collisions within one group is
a common case - see the k8s example alerts;
* supports metrics unregistering on rule updates. Consider the case when one rule
was added or removed from the group, or the whole group was added or removed.
The change depends on https://github.com/VictoriaMetrics/metrics/pull/16
where race condition for Unregister method was fixed.
`external.label` flag supposed to help to distinguish alert or recording rules
source in situations when more than one `vmalert` runs for the same datasource
or AlertManager.
