94 Commits

Author	SHA1	Message	Date
Aliaksandr Valialkin	27afe7bc38	app/vmselect/promql: reduce the number of memory allocations inside getCommonLabelFilters() ... This should improve performance a bit for `q1 op q2` queries	2023-01-15 13:03:23 -08:00
Aliaksandr Valialkin	7067e8206c	app/vmselect/promql: reduce memory allocations at getCommonLabelFilters() function ... Intern tag keys and values there	2023-01-12 01:27:41 -08:00
Aliaksandr Valialkin	31fc29599f	app/vmselect/promql: move the eval function args in parallel query trace outside the loop	2023-01-10 22:23:30 -08:00
Aliaksandr Valialkin	b796a0dc3f	app/vmselect/promql: optimize e1 op e2 when e1 returns an empty result ... Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3349	2022-11-21 16:09:10 +02:00
Aliaksandr Valialkin	b8da90b893	app/vmselect/promql: properly handle zero and negative values for -search.maxMemoryPerQuery ... This is a follow-up for 04a05f161c Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3203	2022-10-12 09:25:17 +03:00
Aliaksandr Valialkin	04a05f161c	app/vmselect: return back the logic for limits the amounts of memory occupied by concurrently executed queries if -search.maxMemoryPerQuery isn't set ... This is needed for preserving backwards compatibility with the previous releases of VictoriaMetrics. Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3203	2022-10-10 21:45:13 +03:00
Aliaksandr Valialkin	5138eaeea0	app/vmselect: allow limiting per-query memory usage via -search.maxMemoryPerQuery command-line flag ... Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3203	2022-10-08 01:08:05 +03:00
Dmytro Kozlov	4415c71a2b	vmselect/{promql, prometheus}: show flag names which user can update in error message (#3049) ... * vmselect/{promql, prometheus}: show flag names which user can update in error message * vmselect/{promql, prometheus}: fix typo	2022-09-06 13:25:59 +03:00
Aliaksandr Valialkin	4076277cf0	app/vmselect/promql: evaluate union() args in parallel in order to increase query performance ... Note that the parallel execution of `union()` args may take more memory and CPU time than the sequential execution if args contain heavy queries, which may load all the available CPU, disk and memory resources and vmselect and vmstorage levels.	2022-09-02 19:46:27 +03:00
Aliaksandr Valialkin	ad11b8d83d	app/vmselect/promql: follow-up after 2d71b4859c ... - Use getScalar() function for obtaining the expected scalar from phi arg - Reduce the error message returned to the user when incorrect phi is passed to histogram_quantiles - Improve the description of this bugfix in the docs/CHANGELOG.md Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3026	2022-08-27 01:35:49 +03:00
Dmytro Kozlov	463ea6897b	vmselect/promql: enable search.maxPointsSubqueryPerTimeseries for sub-queries (#2963) ... * vmselect/promql: enable search.maxPointsPerTimeSeriesSubquery for sub-queries * vmselect/promql: cleanup * vmselect/promql: rename config flag * vmselect/promql: add tests * vmselect/promql: use test object instead of log * vmselect/promql: fix posible panic is subquery has more points. add description * vmselect/promql: update tests descriptions * vmselect/promql: update doInternal validation * vmselect/promql: fix linter * vmselect/promql: fix linter * vmselect/promql: update documentation and release notes * wip - Properly apply -search.maxPointsSubqueryPerTimeseries limit to subqueries. Previously the -search.maxPointsPerTimeseries limit was unexpectedly applied to subqueries if it was smaller than the -search.maxPointsSubqueryPerTimeseries . - Clarify docs for -search.maxPointsSubqueryPerTimeseries command-line flag . - Document -search.maxPointsPerTimeseries and -search.maxPointsSubqueryPerTimeseries flags at https://docs.victoriametrics.com/#resource-usage-limits . - Update docs/CHANGELOG.md . Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/2922 Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>	2022-08-24 15:25:18 +03:00
Aliaksandr Valialkin	4ac79d29ad	app/vmselect: follow-up after 63e0f16062 ... * Explicitly store a pointer to UserReadableError in the error interface. Previously Go automatically converted the value to a pointer before storing in the error interface. * Add Unwrap() method to UserReadableError, so it can be used transparently with the other code, which calls errors.Is() and errors.As(). * Document the change in docs/CHANGELOG.md	2022-08-15 13:50:16 +03:00
Roman Khavronenko	63e0f16062	vmselect: introduce UserReadableError type of error (#2894) ... When read query fails, VM returns rich error message with all the details. While these details might be useful for debugging specific cases, they're usually too verbose for users. Introducing a new error type `UserReadableError` is supposed to allow to return to user only the most important parts of the error trace. This supposed to improve error readability in web interfaces such as VMUI or Grafana. The full error trace is still logged with the full context and can be found in vmselect logs. Signed-off-by: hagen1778 <roman@victoriametrics.com> Signed-off-by: hagen1778 <roman@victoriametrics.com>	2022-08-15 13:38:47 +03:00
Roman Khavronenko	9ccf695d57	vmselect: return correct error for second part of expression (#2893) ... Signed-off-by: hagen1778 <roman@victoriametrics.com>	2022-07-20 16:44:28 +02:00
Aliaksandr Valialkin	cc7d499bbd	app/vmselect/promql: execute q1 and q2 from q1 op q2 in parallel if labels pushdown cannot be applied ... This should improve query performance if VictoriaMetrics has enough resources for processing `q1` and `q2` in parallel. Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/2886	2022-07-19 14:27:48 +03:00
Aliaksandr Valialkin	c2197ad139	app/vmselect/promql: validate function name before evaluating its arguments ... This avoids unneeded evaluation of args for unknown functions	2022-07-12 19:48:26 +03:00
Aliaksandr Valialkin	3e2dd85f7d	all: readability improvements for query traces ... - show dates in human-readable format, e.g. 2022-05-07, instead of a numeric value - limit the maximum length of queries and filters shown in trace messages	2022-06-30 18:20:33 +03:00
Aliaksandr Valialkin	a14188dd8e	app/vmselect: expose additional histograms at /metrics page, which may help get more insights for the query workload ... This commit is based on https://github.com/VictoriaMetrics/VictoriaMetrics/pull/2792	2022-06-28 20:18:13 +03:00
Aliaksandr Valialkin	a43f2d0bc5	app/vmselect/promql: show the number of scanned samples in the query trace	2022-06-28 19:26:17 +03:00
Aliaksandr Valialkin	e578549b8a	app/vmselect: optimize /api/v1/series a bit for time ranges smaller than one day	2022-06-28 13:02:47 +03:00
Aliaksandr Valialkin	a963b2a0aa	all: show timeRange in traces in human-readable format instead of timestamps in milliseconds	2022-06-27 13:45:51 +03:00
Aliaksandr Valialkin	12ac255dae	lib/querytracer: make it easier to use by passing trace context message to New and NewChild ... The context message can be extended by calling Donef. If there is no need to extend the message, then just call Done.	2022-06-08 21:06:52 +03:00
Aliaksandr Valialkin	41958ed5dd	all: add initial support for query tracing ... See https://docs.victoriametrics.com/Single-server-VictoriaMetrics.html#query-tracing Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1403	2022-06-01 02:29:23 +03:00
Aliaksandr Valialkin	6e364e19ef	app/vmselect: add fine-grained limits for the number of returned/scanned time series for various APIs	2022-03-26 11:29:49 +02:00
Aliaksandr Valialkin	0d47c23a03	app/vmselect/promql: reduce the maximum number of label values, which can be propagated from one side of the binary operation to another side of the binary operation from 10K to 1K ... There are user reports that 10K unique values in a single label filter may lead to performance and memory usage issues	2022-02-24 04:05:18 +02:00
Aliaksandr Valialkin	b1f94f7f0e	app/vmselect/promql: return at most one time series from absent_over_time() in the same way as Prometheus does ... See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/2130	2022-02-12 15:45:09 +02:00
Aliaksandr Valialkin	96b7de6736	app/vmselect/promql: clarify comments on why the right side of if and and operators are executed at first	2022-02-03 00:26:14 +02:00
Aliaksandr Valialkin	4b850c2a59	app/vmselect/promql: do not push down filters, which enumerate more than 10k unique values ... Such filters may slow down time series search, so just skip them. This is a follow-up for e7f1ceeb84 Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1827	2022-02-02 23:40:02 +02:00
Aliaksandr Valialkin	2016a2c899	app/vmselect/promql: properly handle foo or bar queries ... Such queries may miss `bar` results after the commit e7f1ceeb84 because common label filters from `foo` could be mistakenly applied to `bar`.	2022-02-01 17:40:51 +02:00
Aliaksandr Valialkin	a8d22e1223	app/vmselect/promql: check for binary operation in case-insensitive manner when deciding which side of the operation to perform the first ... PromQL and MetricsQL operators are case-insensitive	2022-01-31 22:06:15 +02:00
Aliaksandr Valialkin	e7f1ceeb84	app/vmselect/promql: optimize queries, which join on _info metrics. ... Automatically add common filters from one side of binary operation to the other side before sending the query to storage subsystem. See https://grafana.com/blog/2021/08/04/how-to-use-promql-joins-for-more-effective-queries-of-prometheus-metrics-at-scale/ and https://www.robustperception.io/exposing-the-software-version-to-prometheus	2022-01-31 19:32:36 +02:00
Aliaksandr Valialkin	dc7b63a793	app/vmselect/promql: properly keep metric names when optimized path is used for aggregate function calculations ... For example, `sum(rate(...) keep_metric_names) by (__name__)` didn't leave the original metric name because of this issue. This is a follup-up for 1bdc71d917 Udates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/949	2022-01-17 15:30:30 +02:00
Aliaksandr Valialkin	1bdc71d917	app/vmselect/promql: implement keep_metric_names modifier for transform and rollup functions ... Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/949	2022-01-14 04:14:59 +02:00
Aliaksandr Valialkin	f41846d002	app/vmselect/promql: add stale_samples_over_time() function	2022-01-14 01:48:04 +02:00
Aliaksandr Valialkin	c883c15878	app/vmselect/promql: add support for @ modifier ... Add support for `@` modifier in MetricsQL according to https://prometheus.io/docs/prometheus/latest/querying/basics/#modifier Extend the support with the following features: * Allow using `@` modifier everywhere in the query. For example, `sum(foo) @ end()` * Allow using arbitrary expression as `@` modifier. For example, `foo @ (end() - 1h)` returns `foo` value at `end - 1 hour` timestamp on the selected time range `[start ... end]` Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1348	2022-01-13 22:12:06 +02:00
Aliaksandr Valialkin	ff15a752c1	app/vmselect: accept optional extra_filters[] query args for all the supported Prometheus querying APIs ... See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1863	2021-12-06 17:07:09 +02:00
Michael Fuller	cf8c171f85	vmselect: in promql evaluation, return bytes requested when rollup memory limiter is unable to satisfy the request (#1838) ... Co-authored-by: Michael Fuller <mfuller@digitalocean.com>	2021-11-22 13:20:42 +03:00
Aliaksandr Valialkin	9a3d0c43b5	app/vmselect/promql: add quantiles_over_time("phiLabel", phi1, ..., phiN, m[d]) function for calculating multiple quantiles at once	2021-09-17 23:35:10 +03:00
Aliaksandr Valialkin	cdc372bb98	app/vmselect: add -search.noStaleMarkers command-line flag for disabling stale markers handling in queries ... This option allows reducing CPU usage a bit when VictoriaMetrics is used for collecting and processing non-Prometheus data. For example, InfluxDB line protocol, Graphite, OpenTSDB, CSV, etc.	2021-08-18 13:59:02 +03:00
Aliaksandr Valialkin	bd14b0887e	app/vmselect/promql: move common condition to dropStaleNaNs in order to improve code maintainability	2021-08-17 11:01:16 +03:00
Aliaksandr Valialkin	113f0a8a07	app/vmselect/promql: drop staleness marks before calling rollupConfig.Do ... This allows dropping staleness marks only once and then calculate multiple rollup functions on the result. Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1526	2021-08-15 13:21:10 +03:00
Aliaksandr Valialkin	a925d5a3e1	app/vmselect/promql: duration handling improvements in MetricsQL queries ... - Support durations anywhere in MetricsQL queries. E.g. sum_over_time(m[1h])/1h is equivalent to sum_over_time(m[1h])/3600 - Support durations without suffix. E.g. rate(m[300]) is equivalent to rate(m[5m])	2021-07-12 17:16:41 +03:00
Aliaksandr Valialkin	71b72304ae	app/vmselect: improve description for -search.maxPointsPerTimeseries command-line flag	2021-03-22 16:45:34 +02:00
Aliaksandr Valialkin	2dae0a2c47	app/vmselect: add round_digits query arg to /api/v1/query and /api/v1/query_range handlers for limiting the number of decimal digits after the point	2021-03-15 12:36:33 +02:00
Aliaksandr Valialkin	2d33230793	app/vmselect/promql: properly make copies of EvalConfig	2021-02-11 12:41:15 +02:00
Aliaksandr Valialkin	7a3a9421f3	app/vmselect/promql: make a copy of EvalConfig when executing q1 and q2 in parallel for q1 binary_op q2 ... This should prevent from data races if the underlying functions modify EvalConfig contents.	2021-02-10 23:05:55 +02:00
Aliaksandr Valialkin	04faea8b45	app/vmselect: parallelize q1 <binary_op> q2 queries by running q1 and q2 in parallel ... This should reduce query execution times.	2021-02-10 22:59:39 +02:00
Aliaksandr Valialkin	b43b498fd8	app/vmselect: add ability to pass extra_label=<label>=<value> query arg to Prometheus Querying API ... This enforced `{label="value"}` label filter to the query. See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1021	2021-02-01 18:04:17 +02:00
Aliaksandr Valialkin	c164a8d231	app/vmselect/promql: improve documentation for -search.maxPointsPertimeseries command-line flag ... This should reduce incorrect usage and assumptions for this flag. See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1020	2021-01-22 13:00:10 +02:00
Aliaksandr Valialkin	c86286ec1d	app/vmselect/promql: do not ajdust offset value provided in the query ... Previously it could be modified in order to improve response cache hit ratio. This is unneeded, since cache hit ratio should remain good because the query time range should be already aligned to multiple of `step` values. See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/976	2020-12-27 14:09:25 +02:00