mirror of
https://github.com/VictoriaMetrics/VictoriaMetrics.git
synced 2024-11-21 14:44:00 +00:00
app/vmselect/prometheus: do not drop match[] filters if -search.ignoreExtraFiltersAtLabelsAPI flag is set
The `match[]` filter is mandatory at /api/v1/series, so it mustn't be dropped here.
There is no sense in dropping `match[]` filter together with `extra_label` and `extra_filters[]`
at /api/v1/labels and /api/v1/label/.../values if -search.ignoreExtraFiltersAtLabelsAPI commnad-line flag is set,
since:
- the `match[]` filter triggers slow path at these APIs;
- the `extra_label` and `extra_filters[]` filters narrow down the number of matched time series,
so they improve performance comparing to the case when only `match[]` filter is left,
while `extra_label` and `extra_filters[]` filters are dropped.
This is a follow-up for 0b7a23a91d
This commit is contained in:
Aliaksandr Valialkin
parent
e036433b8b
commit
b33b620af6
5 changed files with 36 additions and 31 deletions
15
README.md
15
README.md
|
|
@ -1675,8 +1675,14 @@ By default, VictoriaMetrics is tuned for an optimal resource usage under typical
|
|||
This allows saving CPU and RAM when executing unexpected heavy queries.
|
||||
- `-search.maxConcurrentRequests` limits the number of concurrent requests VictoriaMetrics can process. Bigger number of concurrent requests usually means
|
||||
bigger memory usage. For example, if a single query needs 100 MiB of additional memory during its execution, then 100 concurrent queries may need `100 * 100 MiB = 10 GiB`
|
||||
of additional memory. So it is better to limit the number of concurrent queries, while suspending additional incoming queries if the concurrency limit is reached.
|
||||
VictoriaMetrics provides `-search.maxQueueDuration` command-line flag for limiting the max wait time for suspended queries. See also `-search.maxMemoryPerQuery` command-line flag.
|
||||
of additional memory. So it is better to limit the number of concurrent queries, while pausing additional incoming queries if the concurrency limit is reached.
|
||||
VictoriaMetrics provides `-search.maxQueueDuration` command-line flag for limiting the max wait time for paused queries. See also `-search.maxMemoryPerQuery` command-line flag.
|
||||
- `-search.maxQueueDuration` limits the maximum duration queries may wait for execution when `-search.maxConcurrentRequests` concurrent queries are executed.
|
||||
- `-search.ignoreExtraFiltersAtLabelsAPI` enables ignoring of `match[]`, [`extra_filters[]` and `extra_label`](https://docs.victoriametrics.com/#prometheus-querying-api-enhancements)
|
||||
query args at [/api/v1/labels](https://docs.victoriametrics.com/url-examples/#apiv1labels) and
|
||||
[/api/v1/label/.../values](https://docs.victoriametrics.com/url-examples/#apiv1labelvalues).
|
||||
This may be useful for reducing the load on VictoriaMetrics if the provided extra filters match too many time series.
|
||||
The downside is that the endpoints can return labels and series, which do not match the provided extra filters.
|
||||
- `-search.maxSamplesPerSeries` limits the number of raw samples the query can process per each time series. VictoriaMetrics sequentially processes
|
||||
raw samples per each found time series during the query. It unpacks raw samples on the selected time range per each time series into memory
|
||||
and then applies the given [rollup function](https://docs.victoriametrics.com/MetricsQL.html#rollup-functions). The `-search.maxSamplesPerSeries` command-line flag
|
||||
|
|
@ -1719,11 +1725,6 @@ By default, VictoriaMetrics is tuned for an optimal resource usage under typical
|
|||
when the database contains big number of unique time series because of [high churn rate](https://docs.victoriametrics.com/FAQ.html#what-is-high-churn-rate).
|
||||
In this case it might be useful to set the `-search.maxLabelsAPIDuration` to quite low value in order to limit CPU and memory usage.
|
||||
See also `-search.maxLabelsAPISeries` and `-search.ignoreExtraFiltersAtLabelsAPI`.
|
||||
- `-search.ignoreExtraFiltersAtLabelsAPI` enables ignoring of `match[]`, [`extra_filters[]` and `extra_label`](https://docs.victoriametrics.com/#prometheus-querying-api-enhancements)
|
||||
query args at [/api/v1/labels](https://docs.victoriametrics.com/url-examples/#apiv1labels) and
|
||||
[/api/v1/label/.../values](https://docs.victoriametrics.com/url-examples/#apiv1labelvalues).
|
||||
This may be useful for reducing the load on VictoriaMetrics if the provided extra filters match too many time series.
|
||||
The downside is that the endpoints can return labels and series, which do not match the provided extra filters.
|
||||
- `-search.maxTagValueSuffixesPerSearch` limits the number of entries, which may be returned from `/metrics/find` endpoint. See [Graphite Metrics API usage docs](#graphite-metrics-api-usage).
|
||||
|
||||
See also [resource usage limits at VictoriaMetrics cluster](https://docs.victoriametrics.com/Cluster-VictoriaMetrics.html#resource-usage-limits),
|
||||
|
|
|
|||
|
|
@ -1181,18 +1181,21 @@ func getCommonParamsInternal(r *http.Request, startTime time.Time, requireNonEmp
|
|||
if requireNonEmptyMatch && len(matches) == 0 {
|
||||
return nil, fmt.Errorf("missing `match[]` arg")
|
||||
}
|
||||
|
||||
var filterss [][]storage.TagFilter
|
||||
if !isLabelsAPI || !*ignoreExtraFiltersAtLabelsAPI {
|
||||
tagFilterss, err := getTagFilterssFromMatches(matches)
|
||||
filterss, err := getTagFilterssFromMatches(matches)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
if len(filterss) > 0 || !isLabelsAPI || !*ignoreExtraFiltersAtLabelsAPI {
|
||||
// If matches isn't empty, then there is no sense in ignoring extra filters
|
||||
// even if ignoreExtraLabelsAtLabelsAPI is set, since extra filters won't slow down
|
||||
// the query - they can only improve query performance by reducing the number
|
||||
// of matching series at the storage level.
|
||||
etfs, err := searchutils.GetExtraTagFilters(r)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
filterss = searchutils.JoinTagFilterss(tagFilterss, etfs)
|
||||
filterss = searchutils.JoinTagFilterss(filterss, etfs)
|
||||
}
|
||||
|
||||
cp := &commonParams{
|
||||
|
|
|
|||
|
|
@ -41,6 +41,8 @@ See also [LTS releases](https://docs.victoriametrics.com/lts-releases/).
|
|||
* FEATURE: [stream aggregation](https://docs.victoriametrics.com/stream-aggregation/): expose `vm_streamaggr_flush_timeouts_total` and `vm_streamaggr_dedup_flush_timeouts_total` [counters](https://docs.victoriametrics.com/keyconcepts/#counter) at [`/metrics` page](https://docs.victoriametrics.com/#monitoring), which can be used for detecting flush timeouts for stream aggregation states. Expose also `vm_streamaggr_flush_duration_seconds` and `vm_streamaggr_dedup_flush_duration_seconds` [histograms](https://docs.victoriametrics.com/keyconcepts/#histogram) for monitoring the real flush durations of stream aggregation states.
|
||||
* FEATURE: [vmui](https://docs.victoriametrics.com/#vmui): improve trace display for better visual separation of branches. See [this pull request](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/5926).
|
||||
|
||||
* BUGFIX: do not drop `match[]` filter at [`/api/v1/series`](https://docs.victoriametrics.com/url-examples/#apiv1series) if `-search.ignoreExtraFiltersAtLabelsAPI` command-line flag is set, since this broke the `/api/v1/series` requests.
|
||||
|
||||
## [v1.99.0](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.99.0)
|
||||
|
||||
Released at 2024-03-01
|
||||
|
|
|
|||
|
|
@ -658,11 +658,16 @@ Some workloads may need fine-grained resource usage limits. In these cases the f
|
|||
Bigger number of concurrent requests usually require bigger amounts of memory at both `vmselect` and `vmstorage`.
|
||||
For example, if a single query needs 100 MiB of additional memory during its execution, then 100 concurrent queries
|
||||
may need `100 * 100 MiB = 10 GiB` of additional memory. So it is better to limit the number of concurrent queries,
|
||||
while suspending additional incoming queries if the concurrency limit is reached.
|
||||
`vmselect` and `vmstorage` provides `-search.maxQueueDuration` command-line flag for limiting the maximum wait time for suspended queries.
|
||||
while pausing additional incoming queries if the concurrency limit is reached.
|
||||
`vmselect` and `vmstorage` provides `-search.maxQueueDuration` command-line flag for limiting the maximum wait time for paused queries.
|
||||
See also `-search.maxMemoryPerQuery` command-line flag at `vmselect`.
|
||||
- `-search.maxQueueDuration` at `vmselect` and `vmstorage` limits the maximum duration queries may wait for execution when `-search.maxConcurrentRequests`
|
||||
concurrent queries are executed.
|
||||
- `-search.ignoreExtraFiltersAtLabelsAPI` at `vmselect` enables ignoring of `match[]`, [`extra_filters[]` and `extra_label`](https://docs.victoriametrics.com/#prometheus-querying-api-enhancements)
|
||||
query args at [/api/v1/labels](https://docs.victoriametrics.com/url-examples/#apiv1labels) and
|
||||
[/api/v1/label/.../values](https://docs.victoriametrics.com/url-examples/#apiv1labelvalues).
|
||||
This may be useful for reducing the load on `vmstorage` if the provided extra filters match too many time series.
|
||||
The downside is that the endpoints can return labels and series, which do not match the provided extra filters.
|
||||
- `-search.maxSamplesPerSeries` at `vmselect` limits the number of raw samples the query can process per each time series.
|
||||
`vmselect` processes raw samples sequentially per each found time series during the query. It unpacks raw samples on the selected time range
|
||||
per each time series into memory and then applies the given [rollup function](https://docs.victoriametrics.com/MetricsQL.html#rollup-functions).
|
||||
|
|
@ -709,11 +714,6 @@ Some workloads may need fine-grained resource usage limits. In these cases the f
|
|||
when the database contains big number of unique time series because of [high churn rate](https://docs.victoriametrics.com/FAQ.html#what-is-high-churn-rate).
|
||||
In this case it might be useful to set the `-search.maxLabelsAPIDuration` to quite low value in order to limit CPU and memory usage.
|
||||
See also `-search.maxLabelsAPISeries` and `-search.ignoreExtraFiltersAtLabelsAPI`.
|
||||
- `-search.ignoreExtraFiltersAtLabelsAPI` at `vmselect` enables ignoring of `match[]`, [`extra_filters[]` and `extra_label`](https://docs.victoriametrics.com/#prometheus-querying-api-enhancements)
|
||||
query args at [/api/v1/labels](https://docs.victoriametrics.com/url-examples/#apiv1labels) and
|
||||
[/api/v1/label/.../values](https://docs.victoriametrics.com/url-examples/#apiv1labelvalues).
|
||||
This may be useful for reducing the load on `vmstorage` if the provided extra filters match too many time series.
|
||||
The downside is that the endpoints can return labels and series, which do not match the provided extra filters.
|
||||
- `-storage.maxDailySeries` at `vmstorage` can be used for limiting the number of time series seen per day aka
|
||||
[time series churn rate](https://docs.victoriametrics.com/FAQ.html#what-is-high-churn-rate). See [cardinality limiter docs](#cardinality-limiter).
|
||||
- `-storage.maxHourlySeries` at `vmstorage` can be used for limiting the number of [active time series](https://docs.victoriametrics.com/FAQ.html#what-is-an-active-time-series).
|
||||
|
|
|
|||
|
|
@ -261,9 +261,9 @@ Cluster version of VictoriaMetrics:
|
|||
curl http://<vmselect>:8481/select/0/prometheus/api/v1/labels
|
||||
```
|
||||
|
||||
|
||||
By default, VictoriaMetrics returns labels seen during the last day starting at 00:00 UTC. An arbitrary time range can be set via [`start` and `end` query args](https://docs.victoriametrics.com/#timestamp-formats).
|
||||
The specified `start..end` time range is rounded to day granularity because of performance optimization concerns.
|
||||
By default, VictoriaMetrics returns labels seen during the last day starting at 00:00 UTC because of performance reasons.
|
||||
An arbitrary time range can be set via [`start` and `end` query args](https://docs.victoriametrics.com/#timestamp-formats).
|
||||
The specified `start..end` time range is rounded to UTC day granularity because of performance reasons.
|
||||
|
||||
Additional information:
|
||||
* [Getting label names](https://prometheus.io/docs/prometheus/latest/querying/api/#getting-label-names)
|
||||
|
|
@ -280,16 +280,15 @@ Single-node VictoriaMetrics:
|
|||
curl http://localhost:8428/prometheus/api/v1/label/job/values
|
||||
```
|
||||
|
||||
|
||||
Cluster version of VictoriaMetrics:
|
||||
|
||||
```sh
|
||||
curl http://<vmselect>:8481/select/0/prometheus/api/v1/label/job/values
|
||||
```
|
||||
|
||||
|
||||
By default, VictoriaMetrics returns labels values seen during the last day starting at 00:00 UTC. An arbitrary time range can be set via `start` and `end` query args.
|
||||
The specified `start..end` time range is rounded to day granularity because of performance optimization concerns.
|
||||
By default, VictoriaMetrics returns labels values seen during the last day starting at 00:00 UTC because of performance reasons.
|
||||
An arbitrary time range can be set via `start` and `end` query args.
|
||||
The specified `start..end` time range is rounded to UTC day granularity because of performance reasons.
|
||||
|
||||
Additional information:
|
||||
* [Querying label values](https://prometheus.io/docs/prometheus/latest/querying/api/#querying-label-values)
|
||||
|
|
@ -361,9 +360,9 @@ Cluster version of VictoriaMetrics:
|
|||
curl http://<vmselect>:8481/select/0/prometheus/api/v1/series -d 'match[]=vm_http_request_errors_total'
|
||||
```
|
||||
|
||||
|
||||
By default, VictoriaMetrics returns time series seen during the last day starting at 00:00 UTC. An arbitrary time range can be set via `start` and `end` query args.
|
||||
The specified `start..end` time range is rounded to day granularity because of performance optimization concerns.
|
||||
By default, VictoriaMetrics returns time series seen during the last day starting at 00:00 UTC because of performance reasons.
|
||||
An arbitrary time range can be set via `start` and `end` query args.
|
||||
The specified `start..end` time range is rounded to UTC day granularity because of performance reasons.
|
||||
|
||||
Additional information:
|
||||
* [Finding series by label matchers](https://prometheus.io/docs/prometheus/latest/querying/api/#finding-series-by-label-matchers)
|
||||
|
|
|
|||
Loading…
Reference in a new issue