This pull request updates our fork of Alertmanager to commit 65bdab0, which is based on commit 5658f8c in Prometheus Alertmanager.
It applies the changes from grafana/alerting#155 which removes the overrides for validation of alerts, labels and silences that we had put in place to allow alerts and silences to work for non-Prometheus datasources. However, as this is now supported in Alertmanager with the UTF-8 work, we can use the new upstream functions and remove these overrides.
The compat package is a package in Alertmanager that takes care of backwards compatibility when parsing matchers, validating alerts, labels and silences. It has three modes: classic mode, UTF-8 strict mode, fallback mode. These modes are controlled via compat.InitFromFlags. Grafana initializes the compat package without any feature flags, which is the equivalent of fallback mode. Classic and UTF-8 strict mode are used in Mimir.
While Grafana Managed Alerts have no need for fallback mode, Grafana can still be used as an interface to manage the configurations of Mimir Alertmanagers and view configurations of Prometheus Alertmanager, and those installations might not have migrated or being running on older versions. Such installations behave as if in classic mode, and Grafana must be able to parse their configurations to interact with them for some period of time. As such, Grafana uses fallback mode until we are ready to drop support for outdated installations of Mimir and the Prometheus Alertmanager.
* Alerting: Add metric to check for default AM configurations
* Use a gauge for the config hash
* don't go out of bounds when converting uint64 to float64
* expose metric for config hash
* update metrics after applying config
* Alerting: Add metrics to the remote Alertmanager struct
* rephrase http_requests_failed description
* make linter happy
* remove unnecessary metrics
* extract timed client to separate package
* use histogram collector from dskit
* remove weaveworks dependency
* capture metrics for all requests to the remote Alertmanager (both clients)
* use the timed client in the MimirAuthRoundTripper
* HTTPRequestsDuration -> HTTPRequestDuration, clean up mimir client factory function
* refactor
* less git diff
* gauge for last readiness check in seconds
* initialize LastReadinesCheck to 0, tweak metric names and descriptions
* add counters for sync attempts/errors
* last config sync and last state sync timestamps (gauges)
* change latency metric name
* metric for remote Alertmanager mode
* code review comments
* move label constants to metrics package
* Alerting: Expose metrics for Alertmanager Alerts
In Grafana, the alert evaluation and alert delivery are combined. We're always used a metric named `grafana_alerting_alerts` to get a sense of what are the alerts that are currently firing (these come from the evaluation side) and opted to not map the alertmanager alerts metric directly.
I think it's important that we make a disction between alerts that happen at evaluation vs alerts that are received for delivery by the internal Alertmanager as we have options to skip the delivery of these alerts to the internal alertmanager altogether.
* Alerting: Don't use a separate collection system for metrics
The state package had a metric collection system that ran every 15s updating the values of the metrics - there is a common pattern for this in the Prometheus ecosystem called "collectors".
I have removed the behaviour of using a time-based interval to "set" the metrics in favour of a set of functions as the "value" that get called at scrape time.
* add metrics and tracing to state manager
* propagate tracer to state manager
* add scheduler metrics
* fix backtesting
* add test for state metrics
* remove StateUpdateCount
* update docs
* metrics can be null
* add tracer to new tests
* Create historian metrics and dependency inject
* Record counter for total number of state transitions logged
* Track write failures
* Track current number of active write goroutines
* Record histogram of how long it takes to write history data
* Don't copy the registerer
* Adjust naming of write failures metric
* Introduce WritesTotal to complement WritesFailedTotal
* Measure TransitionsFailedTotal to complement TransitionsTotal
* Rename all to state_history
* Remove redundant Total suffix
* Increment totals all the time, not just on success
* Drop ActiveWriteGoroutines
* Drop PersistDuration in favor of WriteDuration
* Drop unused gauge
* Make writes and writesFailed per org
* Add metric indicating backend and a spot for future metadata
* Drop _batch_ from names and update help
* Add metric for bytes written
* Better pairing of total + failure metric updates
* Few tweaks to wording and naming
* Record info metric during composition
* Create fakeRequester and simple happy path test using it
* Blocking test for the full historian and test for happy path metrics
* Add tests for failure case metrics
* Smoke test for full annotation persistence
* Create test for metrics on annotation persistence, both happy and failing paths
* Address linter complaints
* More linter complaints
* Remove unnecessary whitespace
* Consistency improvements to help texts
* Update tests to match new descs
* Alerting: Add metrics for active receiver and integrations
Introduces metrics that allows us to track the number of configured receivers and integration in the Alertmanager for all orgs.
As a bonus, I realised that the alert reception metrics where not being exported nor collected. This does that too.
* Loki backend and client depend on a requester
* Instrument all requests to loki using weaveworks TimedClient
* Construct collector in metrics package
The `rule_groups_rules` metric is currently defined and computed by `State`.
It makes more sense for this metric to be computed off of the configured rule
set, not based on the rule evaluation state. There could be an edge condition
where a rule does not have a state yet, and so is uncounted.
Additionally, we would like this metric (and others), to have a `rule_group`
label, and this is much easier to achieve if the metric is produced from the
`Scheduler` package.
This change exposes more metrics from the embedded Alertmanager, which are
valuable for troubleshooting Alertmanager operation particularly in HA setups.
```
grafana_alerting_notifications_total
grafana_alerting_notifications_failed_total
grafana_alerting_notification_requests_total
grafana_alerting_notification_requests_failed_total
grafana_alerting_notification_latency_seconds
grafana_alerting_nflog_gc_duration_seconds
grafana_alerting_nflog_snapshot_duration_seconds
grafana_alerting_nflog_snapshot_size_bytes
grafana_alerting_nflog_queries_total
grafana_alerting_nflog_query_errors_total
grafana_alerting_nflog_query_duration_seconds
grafana_alerting_nflog_gossip_messages_propagated_total
grafana_alerting_dispatcher_aggregation_groups
grafana_alerting_dispatcher_alert_processing_duration_seconds
```
Note that `alertmanager_dispatcher_aggregation_group_limit_reached_total` is
explicitly not exposed, as the group limit metrics are not enabled.
* Move ticker files to dedicated package with no changes
* Fix package naming and resolve naming conflicts
* Fix up all existing references to moved objects
* Remove all alerting-specific references from shared util
* Rename TickerMetrics to simply Metrics
* Rename base ticker type to T and rename NewTicker to simply New
Because Summary metrics can not be aggreated, convert them to histograms
so that users with HA deployments can use these metrics.
* Convert metrics registration to promauto.
* Improve help text style.
Signed-off-by: SuperQ <superq@gmail.com>
* pkg/web: store http.Handler internally
* pkg/web: remove injection
Removes any injection code from pkg/web.
It already was no longer functional, as we already only injected into
`http.Handler`, meaning we only inject ctx.Req and ctx.Resp.
Any other types (*Context, *ReqContext) were already accessed using the
http.Request.Context.Value() method.
* *: remove type mappings
Removes any call to the previously removed TypeMapper, as those were
non-functional already.
* pkg/web: remove Context.Invoke
was no longer used outside of pkg/web and also no longer functional
* remove not used code:
- remove offset in ticket because it is not used
- remove unused ticker and scheduler methods
* use duration for interval
* add metrics grafana_alerting_ticker_last_consumed_tick_timestamp_seconds, grafana_alerting_ticker_next_tick_timestamp_seconds, grafana_alerting_ticker_interval_seconds
* pass url parameters through context.Context
* fix url param names without colon prefix
* change context params to vars
* replace url vars in tests using new api
* rename vars to params
* add some comments
* rename seturlvars to seturlparams
* Alerting: Refactor & fix unified alerting metrics structure
Fixes and refactors the metrics structure we have for the ngalert service. Now, each component has its own metric struct that includes the JUST the metrics it uses. Additionally, I have fixed the configuration metrics and added new metrics to determine if we have discovered and started all the necessary configurations of an instance.
This allows us to alert on `grafana_alerting_discovered_configurations - grafana_alerting_active_configurations != 0` to know whether an alertmanager instance did not start successfully.
Fixes#30144
Co-authored-by: dsotirakis <sotirakis.dim@gmail.com>
Co-authored-by: Marcus Efraimsson <marcus.efraimsson@gmail.com>
Co-authored-by: Ida Furjesova <ida.furjesova@grafana.com>
Co-authored-by: Jack Westbrook <jack.westbrook@gmail.com>
Co-authored-by: Will Browne <wbrowne@users.noreply.github.com>
Co-authored-by: Leon Sorokin <leeoniya@gmail.com>
Co-authored-by: Andrej Ocenas <mr.ocenas@gmail.com>
Co-authored-by: spinillos <selenepinillos@gmail.com>
Co-authored-by: Karl Persson <kalle.persson@grafana.com>
Co-authored-by: Leonard Gram <leo@xlson.com>
Introduces org-level isolation for the Alertmanager and its components.
Silences, Alerts and Contact points are not separated by org and are not shared between them.
Co-authored with @davidmparrott and @papagian
