Commit Graph

550 Commits

Author SHA1 Message Date
Aliaksandr Valialkin
202d8e2c40
docs: update -help output after 61d9df4c36
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/834
2024-02-08 14:50:56 +02:00
Aliaksandr Valialkin
b18e608016
app/vmselect: add ability to reset rollup result cache on startup by passing -search.resetRollupResultCacheOnStartup command-line flag
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/834
2024-02-08 14:42:15 +02:00
Aliaksandr Valialkin
def5573f92
app/vmselect/promql: properly handle precision errors in rollup functions
changes(), increases_over_time() and resets() shouldn't take into account
value changes, which may occur because of precision errors.

The maximum guaranteed precision for raw samples stored in VictoriaMetrics is 12 decimal digits.
So do not count relative changes for values if they are smaller than 1e-12 comparing to the value.

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/767
2024-02-08 02:33:17 +02:00
Aliaksandr Valialkin
2033fe4caf
app/vmselect/promql: really keep metric names when keep_metric_names modifier is applied to binary operator
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5556
2024-01-31 02:33:06 +02:00
Roman Khavronenko
02e609b141
app/vmselect: set proper timestamp for cached instant responses ()
* app/vmselect: set proper timestamp for cached instant responses

The change updates `getSumInstantValues` to prefer timestamp
from the most recent results. Before, timestamp from cached series
was used.

The old behavior had negative impact on recording rules as they
were getting responses with shifted timestamps in past.
Subsequent recording or alerting rules fetching results of these
recording rules could get no result due to staleness interval.

https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5659
Signed-off-by: hagen1778 <roman@victoriametrics.com>

* wip

---------

Signed-off-by: hagen1778 <roman@victoriametrics.com>
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>
2024-01-30 22:20:16 +02:00
Aliaksandr Valialkin
e20cbfcbc3
app/vmselect/promql: remove superflouos memory allocations at aggrPrepareSeries()
While at it, also remove unneeded map lookup
2024-01-23 02:29:14 +02:00
Aliaksandr Valialkin
b7cc1af3eb
app/vmselect/promql/aggr_incremental.go: eliminate unnecessary memory allocation in incrementalAggrFuncContext.updateTimeseries 2024-01-23 02:29:13 +02:00
Aliaksandr Valialkin
d4a1a28543
app/vmselect: handle negative time range start in a generic manner inside NewSearchQuery()
This is a follow-up for cf03e11d89

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5553
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/pull/5630
2024-01-22 01:39:27 +02:00
Hui Wang
e086ef16da
app/vmselect/promql: properly handle possible negative results caused… ()
* app/vmselect/promql: properly handle possible negative results caused by float operations precision error in rollup functions like rate() or increase()

* fix test
2024-01-22 01:04:50 +02:00
Aliaksandr Valialkin
84f11a9e6d
app/vmselect/promql: simplify the code after 388d020b7c
Add a test, which verifies the correct sorting of float64 slices with NaNs.

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5506
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/pull/5509
2024-01-16 22:35:51 +02:00
Aliaksandr Valialkin
6ba2fd3312
app/vmselect/promql: follow-up for ce4f26db02
- Document the bugfix at docs/CHANGELOG.md
- Filter out NaN values before sorting as suggested at https://github.com/VictoriaMetrics/VictoriaMetrics/pull/5509#discussion_r1447369218
- Revert unrelated changes in lib/filestream and lib/fs
- Use simpler test at app/vmselect/promql/exec_test.go

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/pull/5509
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5506
2024-01-16 22:13:13 +02:00
Zongyang
cb37df5723
FIX bottomk doesn't return any data when there are no time range overlap between timeseries ()
* FIX sort order in bottomk

* Add lessWithNaNsReversed for bottomk

* Add ut for TopK

* Move lt from loop

* FIX lint

* FIX lint

* FIX lint

* Mod log format

---------

Co-authored-by: xiaozongyang <xiaozngyang@kanyun.com>
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>
2024-01-16 22:12:49 +02:00
Aliaksandr Valialkin
015c0a4d1a
app/vmselect/promql: consistently sort results of a or b query
Previously the order of results returned from `a or b` query could change with each request
because the sorting for such query has been disabled in order to satisfy
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/4763 .

This commit executes `a or b` query as `sortByMetricName(a) or sortByMetricName(b)`.
This makes the order of returned time series consistent across requests,
while maintaining the requirement from https://github.com/VictoriaMetrics/VictoriaMetrics/issues/4763 ,
e.g. `b` results are consistently put after `a` results.

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5393
2024-01-16 22:12:15 +02:00
Roman Khavronenko
4837616df6
app/vmselect: drop rollupDefault function as duplicate ()
* app/vmselect: drop `rollupDefault` function as duplicate

It is unclear why there are two identical fns `rollupDefault`
and `rollupDistinct`. Dropping one of them.

Signed-off-by: hagen1778 <roman@victoriametrics.com>

* Update app/vmselect/promql/rollup.go

* Update app/vmselect/promql/rollup.go

---------

Signed-off-by: hagen1778 <roman@victoriametrics.com>
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>
2023-12-21 11:23:20 +02:00
Anton Tykhyy
51af1dfff7
Fix sum(aggr_over_time) 'got 1 args' error () ()
app/vmselect/promql/eval.go:evalAggrFunc shunts evaluation
of AggrFuncExpr over rollupFunc over MetricsExpr to an optimized
path. tryGetArgRollupFuncWithMetricExpr() checks whether expression
can be shunted, but it mangles the AggrFuncExpr when the aggregation
function has more than one argument. This results in queries like
`sum(aggr_over_time("avg_over_time",m))` failing with error message
'expecting at least 2 args to "aggr_over_time"; got 1 args' while
the analogous query `sum(avg_over_time(m))` executes successfully.
This fix removes the unnecessary mangling.

Signed-off-by: Anton Tykhyy <atykhyy@gmail.com>
2023-12-14 12:49:01 +02:00
Aliaksandr Valialkin
3d6517b05e
app/vmselect: add -search.maxResponseSeries command-line flag for limiting the number of time series a single response can return
This limit can be used for preventing from high memory usage at Grafana when the response returns too many series.

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5372
2023-12-10 00:54:32 +02:00
Aliaksandr Valialkin
509339bf63
app/vmselect: properly adjust the lower bound for the time range where raw samples must be selected for default_rollup() function
Previously the lower bound could be too small, which could result in missing values at the beginning of the graph
for default_rollup() function. This function is automatically applied to all the series selectors if they aren't
explicitly wrapped into a rollup function - see https://docs.victoriametrics.com/MetricsQL.html#implicit-query-conversions

While at it, properly take into account `-search.minStalenessInterval` command-line flag when adjusting
the lower bound for the selected time range.

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5388
2023-12-06 14:46:18 +02:00
luckyxiaoqiang
8ce82c5400
app/vmselect/promql: add day_of_year() function ()
Co-authored-by: dingxiaoqiang <dingxiaoqiang@bytedance.com>
Co-authored-by: Roman Khavronenko <roman@victoriametrics.com>
(cherry picked from commit d7897e0d70)
2023-11-28 12:49:48 +01:00
Aliaksandr Valialkin
5492ccf0d5
app/vmselect/promql: reduce the number of memory allocations inside copyTimeseriesShallow()
Previously the number of memory allocations inside copyTimeseriesShallow() was equal to 1+len(tss)
Reduce this number to 2 by pre-allocating a slice of timeseries structs with len(tss) length.
2023-11-17 15:41:38 +01:00
Aliaksandr Valialkin
633ec37022
app/vmselect/promql: typo fix after 7ca8ebef20
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5332
2023-11-16 17:01:19 +01:00
Aliaksandr Valialkin
7ca8ebef20
app/vmselect/promql: properly handle duplicate series when merging cached results with the results obtained from the database
evalRollupFuncNoCache() may return time series with identical labels (aka duplicate series)
when performing queries satisfying all the following conditions:

- It must select time series with multiple metric names. For example, {__name__=~"foo|bar"}
- The series selector must be wrapped into rollup function, which drops metric names. For example, rate({__name__=~"foo|bar"})
- The rollup function must be wrapped into aggregate function, which has no streaming optimization.
  For example, quantile(0.9, rate({__name__=~"foo|bar"})

In this case VictoriaMetrics shouldn't return `cannot merge series: duplicate series found` error.
Instead, it should fall back to query execution with disabled cache.

Also properly store the merged results. Previously they were incorrectly stored because of a typo
introduced in the commit 41a0fdaf39

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5332
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/pull/5337
2023-11-16 16:16:17 +01:00
Aliaksandr Valialkin
2f885d8e57
app/vmselect/promql: typo fixes after 7cf7740d18 2023-11-14 03:34:25 +01:00
Aliaksandr Valialkin
9ff1ee333f
app/vmselect/promql: properly handle instant query optimization conrner cases for min_over_time() and max_over_time()
- If min_over_time(m[offset] @ timestamp) <= min_over_time(m[offset] @ (timestamp-window)),
  then the optimization can be applied.

- If max_over_time(m[offset] @ timestamp) >= max_over_time(m[offset] @ (timestamp-window)),
  then the optimization can be applied.
2023-11-14 02:58:18 +01:00
Noah Labrecque
fbb572a180
fix: apply correct bounds to sf and tf () 2023-11-14 01:19:47 +01:00
Aliaksandr Valialkin
a45cbc101f
all: cleanup: remove // +build ... lines, since they are no longer needed after Go1.17, and the minimum supported Go version for VictoriaMetrics source code is Go1.20 2023-11-13 19:15:42 +01:00
Aliaksandr Valialkin
d9ecc3f6d7
lib/logger: add -loggerMaxArgLen command-line flag for fine-tuning the maximum length of logged args 2023-11-13 09:43:49 +01:00
Aliaksandr Valialkin
c916294b61
app/vmselect/promql: optimize instant queries with min_over_time() and max_over_time() rollup functions
This is a follow-up for 41a0fdaf39
2023-11-13 09:43:18 +01:00
Aliaksandr Valialkin
bf01a97f17
docs/CHANGELOG.md: update the description of the optimization for SLO/SLI-like queries according to latest changes
See commits 4497a08e3d and 92826b0b4a
2023-11-02 20:09:22 +01:00
Aliaksandr Valialkin
7fc5178a4b
app/vmselect/promql: add missing trace message in rollupResultCache.GetSeries() 2023-11-02 09:17:13 +01:00
Aliaksandr Valialkin
ece7024f11
app/vmselect/promql: reduce the minimum lookbehind window for enabling SLO/SLI optimizations from 24 hours to 6 hours
This reduction is based on production testing.

Also expose -search.minWindowForInstantRollupOptimization command-line flag, so users could fine-tune this arg for their needs
2023-11-01 20:19:19 +01:00
Aliaksandr Valialkin
6a98f9df54
app/vmui: show query execution duration in the header of query input field
This should simplify the process of query optimization
2023-11-01 16:46:42 +01:00
Aliaksandr Valialkin
c5e3b11762
app/vmselect/promql: apply SLO-like optimization to all the count_*_over_time() functions
This is a follow-up for 41a0fdaf39
2023-11-01 09:58:50 +01:00
Aliaksandr Valialkin
b96d55e1e4
app/vmselect/promql: typo fix, which could lead to panic during range query execution
The panic is:

  BUG: unexpected values after merging new values

This is a follow-up for 41a0fdaf39
2023-11-01 09:58:50 +01:00
Aliaksandr Valialkin
7b7ad44e84
app/vmselect/promql: properly calculate rollup result if lookbehind window isn't set
This is a follow-up for 41a0fdaf39
2023-10-31 22:23:04 +01:00
Aliaksandr Valialkin
744f8c3fe7
app/vmselect/promql: add outliers_iqr(q) and outlier_iqr_over_time(m[d]) functions
These functions allow detecting anomalies in series and samples using Interquartile range method.
See Outliers section at https://en.wikipedia.org/wiki/Interquartile_range for more details.
2023-10-31 22:14:14 +01:00
Aliaksandr Valialkin
9661918bb4
app/vmselect/promql: optimize repeated SLI-like instant queries with lookbehind windows >= 1d
Repeated instant queries with long lookbehind windows, which contain one of the following rollup functions,
are optimized via partial result caching:

- sum_over_time()
- count_over_time()
- avg_over_time()
- increase()
- rate()

The basic idea of optimization is to calculate

  rf(m[d] @ t)

as

  rf(m[offset] @ t) + rf(m[d] @ (t-offset)) - rf(m[offset] @ (t-d))

where rf(m[d] @ (t-offset)) is cached query result, which was calculated previously

The offset may be in the range of up to 1 hour.
2023-10-31 20:08:38 +01:00
Aliaksandr Valialkin
9ba007a636
app/vmselect/promql: wrap too long line after a950873fff 2023-10-31 19:11:05 +01:00
Roman Khavronenko
9d8f93050c
app/vmselect: expose vm_memory_intensive_queries_total counter metric ()
The new metric gets increased each time `-search.logQueryMemoryUsage` memory limit
is exceeded by a query. This metric should help to identify expensive and heavy queries
without inspecting the logs.

Signed-off-by: hagen1778 <roman@victoriametrics.com>
2023-10-31 19:02:22 +01:00
Aliaksandr Valialkin
627a4e9330
app/vmselect/promql: add labels_equal(q, "label1", "label2", ...) function
This function returns q series, which have identical values for the listed labels
"label1", "label2", ...

See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5148
2023-10-16 21:51:13 +02:00
Aliaksandr Valialkin
b2f9b9d634
app/vmselect/promql: add drop_empty_series() function for dropping empty series before performing additional calculations
This can be useful in the following queries:

   drop_empty_series(temperature <= 30) default 40

This query drops temperature series with all the values bigger than 30 on the selected time range,
while replacing gaps in the remaining series with 40.

The query without drop_empty_series:

  (temperature <= 30) default 40

would leave all the temperature series with all the values bigger than 30 on the selected time range,
and replace all their values with 40. This is not what could be epxected in some cases
like here - https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5071
2023-10-16 20:59:21 +02:00
Aliaksandr Valialkin
4278b00a66
app/vmselect/promql: do not use unsafe conversion from bytes slice to string when storing a value by map key
The assigned map key shouldn't change over time, otherwise the map won't work properly.

This is a follow-up for 1f91f22b5f
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5087
2023-10-16 13:55:40 +02:00
Nikolay
4a50e9400c
app/vmselect: reduce lock contention for heavy aggregation requests ()
reduce lock contention for heavy aggregation requests
previously lock contetion may happen on machine with big number of CPU due to enabled string interning. sync.Map was a choke point for all aggregation requests.
Now instead of interning, new string is created. It may increase CPU and memory usage for some cases.
https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5087
2023-10-10 13:44:02 +02:00
Aliaksandr Valialkin
5fd79f47f1
app/vmselect/promql: follow-up for 896c85a4a4
- Clarify the description of the change at docs/CHANGELOG.md
- Make sure that bitmap_*(X, NaN) returns NaN

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/4996
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/pull/5021
2023-10-02 21:07:46 +02:00
Dmytro Kozlov
90b189dab8
app/vmselect: fix bitmap_*() functions behavior ()
Related issue: https://github.com/VictoriaMetrics/VictoriaMetrics/issues/4996

Signed-off-by: dmitryk-dk d.kozlov@victoriametrics.com

Signed-off-by: dmitryk-dk d.kozlov@victoriametrics.com
Co-authored-by: Nikolay <nik@victoriametrics.com>
2023-10-02 20:13:27 +02:00
Aliaksandr Valialkin
538dc6058d
app/vmselect/promql: run make fmt after 3b9605dba5 2023-09-25 16:15:58 +02:00
Aliaksandr Valialkin
b43ff80d21
app/vmselect/promql: do not sort q1 or q2 results
This makes sure that `q2` series are returned after `q1` series in the same way as Prometheus does

See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/4763
2023-09-25 16:15:02 +02:00
Aliaksandr Valialkin
c954019e43
app/vmselect/promql: completely substitute median_over_time() WITH template with regular median_over_time() rollup function
This is a follow-up for 34d7a670d0

Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5034
2023-09-25 15:31:25 +02:00
Zakhar Bessarab
fd6ca57c14
app/vmselect/promql: add implementation of median_over_time for rollup functions list ()
`median_over_time` is handled by predefined WITH template in MetricsQL library which translates it to `quantile_over_time(0.5)`
This makes it impossble to use `median_over_time` as a usual rollup function for `aggr_over_time`.

See: https://github.com/VictoriaMetrics/VictoriaMetrics/issues/5034

Signed-off-by: Zakhar Bessarab <z.bessarab@victoriametrics.com>
2023-09-25 15:31:25 +02:00
Dmytro Kozlov
5477b52991
vmagent: add validation of MetricsQL functions ()
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>
2023-09-15 13:16:22 +02:00
Aliaksandr Valialkin
d8afd7fe98
Makefile: update golangci-lint from v1.51.2 to v1.54.2
See https://github.com/golangci/golangci-lint/releases/tag/v1.54.2
2023-09-01 10:25:49 +02:00