1327 Commits

Author	SHA1	Message	Date
Bryan Boreham	7512d13e00	Merge pull request #17002 from bboreham/reduce-promql-benchmark ... [TESTS] PromQL: shrink work done in BenchmarkRangeQuery	2025-08-06 12:07:49 +01:00
Bartlomiej Plotka	5df982538f	Merge pull request #16994 from mmorel-35/unused-parameters ... chore: enable unused-receiver rule from revive	2025-08-06 10:46:28 +01:00
Bartlomiej Plotka	3fc87f40e8	Merge pull request #16998 from mmorel-35/govet/enable-all ... chore: enable-all rule from govet	2025-08-06 10:44:44 +01:00
Bryan Boreham	31eb3b286d	[TESTS] PromQL: shrink work done in BenchmarkRangeQuery ... Previously, BenchmarkRangeQuery would run each case with three data sizes (1, 10, 100) and three range lengths (1, 100, 1000) for nine variations. With 36 cases, running with `-count=6` to get dependable results, would take 40-50 minutes in total. This PR removes the middle option in both dimensions, thus shrinking to four variations and about 20 minutes to run everything. Signed-off-by: Bryan Boreham <bjboreham@gmail.com>	2025-08-05 14:21:23 +01:00
George Krajcsovits	457a2381f9	fix(parser): revert pr 16041 and 16754 (#16996) ... * fix(parser): revert pr 16041 and 16754 Fix for https://github.com/prometheus/prometheus/issues/16053 turned out to be too complicated and lead to more errors than it solved. See https://github.com/prometheus/prometheus/pull/16875. * add TODOs to wrong end positions * restore test and fix initialization bug Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>	2025-08-05 10:48:08 +02:00
Matthieu MOREL	0484a6d547	chore: enable-all rule from govet ... Signed-off-by: Matthieu MOREL <matthieu.morel35@gmail.com>	2025-08-04 18:06:36 +02:00
Matthieu MOREL	cef219c31c	chore: enable unused-receiver rule from revive ... Signed-off-by: Matthieu MOREL <matthieu.morel35@gmail.com>	2025-08-04 09:43:33 +00:00
Bryan Boreham	e9cf792c47	Merge pull request #16895 from bboreham/noinline-kahan ... [BUGFIX] PromQL: Noinline kahanSumInc, to reduce numerical errors	2025-07-28 15:55:19 +01:00
Carrie Edwards	44b0fbba1e	No info annotation for rate/increase when type is histogram (#16915) ... Signed-off-by: Carrie Edwards <edwrdscarrie@gmail.com>	2025-07-24 10:11:37 -07:00
Darshan Chaudhary	9b00c296a8	[PERF] PromQL: Pass concrete types to promql functions (#16797) ... Currently, the promql functions take the interface slice []parser.Value as an argument, which is being implemented by the conrete types Vector, Matrix etc. This PR replaces the interface with the concrete types, resulting in improved performance. The inspiration for this PR came from #16698 which does this for binops. I extended the idea to all promql functions Signed-off-by: darshanime <deathbullet@gmail.com> * pass single Matrix Signed-off-by: darshanime <deathbullet@gmail.com> --------- Signed-off-by: darshanime <deathbullet@gmail.com>	2025-07-23 17:30:03 +01:00
Harry John	b09cf6be8d	fix(promql): Ensure native histogram values copied in subqueries (#16879) ... Signed-off-by: 🌲 Harry 🌊 John 🏔 <johrry@amazon.com>	2025-07-21 08:54:54 +02:00
Bryan Boreham	358499f868	[BUGFIX] PromQL: Noinline kahanSum, to avoid compiler optimisations ... Observed on go1.24.4 darwin/arm64, the compensation variable is not correctly calculated. Note you may have to run the tests several times to see it, to get the right ordering of series. Signed-off-by: Bryan Boreham <bjboreham@gmail.com>	2025-07-19 15:19:29 +01:00
George Krajcsovits	4662722a8e	chore(promql): improve doc string in promql/durations.go (#16863) ... Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>	2025-07-11 12:07:54 +02:00
Björn Rabenstein	b7f984d6d2	Merge pull request #16585 from kapillamba4/fix/16393-strict ... Convert PromQL tests to new syntax via basic migration mode	2025-07-10 15:45:38 +02:00
Björn Rabenstein	eb3ea163fa	promqltest: add tests for `histogram_count(increase(...))` (#16854) ... As `histogram_count` is playing tricks to improve performance, we better make sure that the limitation of extrapolation below zero still works as expected. Signed-off-by: beorn7 <beorn@grafana.com>	2025-07-10 15:44:02 +02:00
Björn Rabenstein	362141370d	Merge pull request #16828 from prometheus/beorn7/histogram2 ... promql(histograms): scale a histogram the same as the count	2025-07-10 13:26:15 +02:00
beorn7	107e4a00c3	promqltest: Test NaN sample values for quantile aggregator ... Signed-off-by: beorn7 <beorn@grafana.com>	2025-07-09 13:38:19 +02:00
liangmulu	b1a7df2c0c	chore: fix some minor issues in comments ... Signed-off-by: liangmulu <liangmulu@outlook.com>	2025-07-09 18:05:41 +08:00
Kapil Lamba	df0e034314	address code review comments ... Signed-off-by: Kapil Lamba <kapillamba4@gmail.com>	2025-07-09 07:25:31 +05:30
Björn Rabenstein	dbee82267a	Merge pull request #16725 from MichaHoffmann/mhoffmann/fix-topk-nan-arg-error-on-nonexisting-series ... promql: fix topk error on NaN argument for non-existing series	2025-07-08 19:42:20 +02:00
beorn7	bcf7a822a0	promql: Prevent extrapolation below zero for histogram count ... This deals with the count field of native histograms in the same way as with simple float counters. It then scale the whole histogram with the same factor as it has scaled the count. This will still allow individual buckets to get extrapolated below zero, but maybe that is fine. This implements approach (2) as described in https://github.com/prometheus/prometheus/issues/15976#issuecomment-3032095158 Signed-off-by: beorn7 <beorn@grafana.com>	2025-07-08 19:01:31 +02:00
Michael Hoffmann	44ee5e2ad6	promql: fix topk error on NaN argument for non-existing series ... Signed-off-by: Michael Hoffmann <mhoffmann@cloudflare.com>	2025-07-07 06:19:39 +00:00
beorn7	c0a13223e7	promql: add tests to demonstrate extrapolation below zero ... This shows how float counters cannot go below zero when extrapolationg for rate/increase, and how histograms do not have that protection yet, leading to an overestimation of the rate/increase. This also demonstrates edge cases where the count extrapolation does not need to be limited, but an individual bucket still goes below zero. Signed-off-by: beorn7 <beorn@grafana.com>	2025-07-06 23:42:55 +02:00
Julien Pivotto	984c8de0da	PromQL: Fix printing +min() ... Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>	2025-07-02 11:17:17 +02:00
Julien Pivotto	3af0bdee68	PromQL: min/max/step: add more tests ... Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>	2025-07-02 11:17:17 +02:00
Julien Pivotto	ee7d5158a7	Add step(), min(a,b) and max(a,b) in promql duration expressions ... step() is a new keyword introduced to represent the query step width in duration expressions. min(a,b) and max(a,b) return the min and max from two duration expressions. Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>	2025-07-02 11:17:17 +02:00
Björn Rabenstein	9e73fb43b3	Merge pull request #16773 from prometheus/beorn7/promql ... promql: Re-introduce direct mean calculation	2025-06-27 14:57:12 +02:00
beorn7	ce809e625f	promql: Re-introduce direct mean calculation for better accuracy ... This commit brings back direct mean calculation (for `avg` and `avg_over_time`) but isn't an outright revert of #16569. It keeps the improved incremental mean calculation and features generally a bit cleaner code than before. Also, this commit... - ...updates the lengthy comment explaining the whole situation and trade-offs. - ...divides the running sum and the Kahan compensation term separately (in direct mean calculation) to avoid the (unlikely) possibility that sum and Kahan compensation together ovorflow float64. - ...uncomments the tests that should now work again on darwin/arm64. - ...uncomments the test that should now reliably yield the (inaccurate) value 0 on all hardware platforms. Also, the test description has been updated accordingly. - ...adds avg_over_time tests for zero and one sample in the range. Signed-off-by: beorn7 <beorn@grafana.com>	2025-06-27 14:34:46 +02:00
beorn7	f71daa7977	promql: Remove falsified comment from test ... The test in question actually worked fine even before #16569. The finding reported in the comment has turned out to be caused by something else. Signed-off-by: beorn7 <beorn@grafana.com>	2025-06-27 14:34:46 +02:00
beorn7	2b3fc1f115	promql: Add test cases for direct mean calculation ... These demonstrate that direct mean calculation has some merits after all. Signed-off-by: beorn7 <beorn@grafana.com>	2025-06-27 14:34:46 +02:00
George Krajcsovits	5b7ff92d95	fix(promql): histogram_quantile and histogram_fraction NaN observed in native histogram (#16724) ... * fix(promql): histogram_quantile NaN observed in native histogram Fixes: #16578 See the issue for detailed explanation. When a histogram had only NaN observations and no normal observations, we returned 0 from the quantile, which is completely wrong. If there were normal observations but we went over them, we returned the upper bound of the existing buckets, however that contradicts expectations on histogram_fraction. Now we return NaN if the quantile is calculated to be over all normal observations, falling into NaNs (in a virtual +Inf bucket). We also return info level annotations if we see any NaN observations. The annotation calls out if we returned NaN or even if we took the virtual +Inf bucket into account. Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com> * fix(promql): histogram_fraction NaN observed in native histogram Fixes: #16580 According to the specification we should not take NaN observations into account when calculating the fraction. This commit fixes that and adds an info level annotation to let the user know about this. Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>	2025-06-25 13:37:43 +02:00
Björn Rabenstein	8ed37d3c09	Merge pull request #16744 from harry671003/fix_last_over_time ... PromQL: Fix native histogram last_over_time	2025-06-23 00:56:08 +02:00
Kapil Lamba	030835560b	promqltest: add/remove explicit annotation expectations to relevant tests ... Signed-off-by: Kapil Lamba <kapillamba4@gmail.com>	2025-06-21 18:46:50 +05:30
Kapil Lamba	658e9c5155	promqltest: migrate test files to new syntax using basic mode ... Signed-off-by: Kapil Lamba <kapillamba4@gmail.com>	2025-06-21 18:46:33 +05:30
Julien	8f75da3fd6	Merge pull request #16754 from simonpasquier/fix-promql-panic ... fix: prevent invalid array access in aggregate expression	2025-06-20 13:53:59 +02:00
Simon Pasquier	71217a6e43	fix: prevent invalid array access in aggregate expression ... This commit fixes the evaluation of invalid expressions like `sum(rate(`. Before that, it would trigger a panic in the PromQL engine because it tried to access an index which is out of range. The bug was probably introduced by 06d0b063ea. Signed-off-by: Simon Pasquier <spasquie@redhat.com>	2025-06-19 15:11:38 +02:00
Kapil Lamba	69906bb4f5	Add script for converting PromQL tests to new syntax format (#16562) ... Signed-off-by: Kapil Lamba <kapillamba4@gmail.com> Co-authored-by: Neeraj Gartia <80708727+NeerajGartia21@users.noreply.github.com>	2025-06-18 23:57:39 +02:00
Charles Korn	8d9dfa075d	promql: reuse `histogramStatsIterator` where possible, and expose it for other implementations to use (#16686) ... * Expose type * Add `Reset` method --------- Signed-off-by: Charles Korn <charles.korn@grafana.com>	2025-06-18 23:53:46 +02:00
Björn Rabenstein	f69c83f5f1	Merge pull request #16702 from bboreham/simpler-detect ... [Refactor] PromQL: Simplify detectHistogramStatsDecoding	2025-06-18 13:02:55 +02:00
🌲 Harry 🌊 John 🏔	59bf3d442b	PromQL: Fix native histogram last_over_time with offset ... Signed-off-by: 🌲 Harry 🌊 John 🏔 <johrry@amazon.com>	2025-06-17 10:49:06 -07:00
Michael Hoffmann	ec98e7f069	promql: fix ts_of_last_over_time for histogram vectors ... This PR fixes a bug in ts_of_last_over_time where the float samples where used when computing the last timestamp of the histogram samples. Signed-off-by: Michael Hoffmann <mhoffmann@cloudflare.com>	2025-06-16 13:09:49 +00:00
Julien	998840c496	Merge pull request #16713 from roidelapluie/fix-pow-dur-expr ... promql: Fix unary operator precedence in duration expressions	2025-06-16 11:13:34 +02:00
Julien Pivotto	6d56e77e0e	promql: Fix unary operator precedence in duration expressions ... Split unary operator handling in duration expressions into two specific cases to fix precedence conflicts: - Handle unary operators with number literals directly - Handle unary operators with parenthesized expressions separately This prevents unary minus from incorrectly binding to subsequent operators in expressions like `foo offset -1^1`, ensuring it parses as `(foo offset -1) ^ 1` rather than `foo offset (-1^1)`. Fixes #16711 Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>	2025-06-13 13:40:50 +02:00
Michael Hoffmann	a5fa9431d8	promql: add ts_of_(max,min,last)_over_time functions ... This commit adds the ts_of_(max,min,last)_over_time functions behind the experimental feature flag. Signed-off-by: Michael Hoffmann <mhoffmann@cloudflare.com>	2025-06-12 15:03:28 +00:00
Innokentii Konstantinov	94d5e0f41c	engine: fix PossibleNonCounterInfo annotation for rate and increase. (#16718) ... * Comment exposed field Signed-off-by: Innokentii Konstantinov <innokenty.konstantinov@grafana.com> * Remove invalid test Signed-off-by: Innokentii Konstantinov <innokenty.konstantinov@grafana.com> * Fix PossibleNonCounterInfo annotation. Signed-off-by: Innokentii Konstantinov <innokenty.konstantinov@grafana.com> * lint Signed-off-by: Innokentii Konstantinov <innokenty.konstantinov@grafana.com> --------- Signed-off-by: Innokentii Konstantinov <innokenty.konstantinov@grafana.com>	2025-06-11 14:09:57 +01:00
beorn7	68b1bcc648	promql: typo fix in comment ... Signed-off-by: beorn7 <beorn@grafana.com>	2025-06-10 22:06:04 +02:00
beorn7	c3da460474	promql: Deactivate three failing tests for the time being ... These tests fail on darwin/arm64. One is expected, because the test demonstrates the limits of the numerical accuracy of our methods, and different inaccurate outcomes on different hardware are expected. The other two are mysterious at the moment, see https://github.com/prometheus/prometheus/issues/16714 for detailed discussion and debugging. Signed-off-by: beorn7 <beorn@grafana.com>	2025-06-10 21:48:52 +02:00
Björn Rabenstein	8fc1750bcc	Merge pull request #16569 from prometheus/beorn7/promql2 ... promql: Simplify avg aggregation and avg_over_time	2025-06-07 00:16:20 +02:00
beorn7	64431830b8	Add more avg_over_time test cases with extreme values ... These tests were initially created by @crush-on-anechka. I modified them slightly. Signed-off-by: beorn7 <beorn@grafana.com>	2025-06-06 23:50:08 +02:00
Bryan Boreham	91eab63d38	[Refactor] PromQL: Simplify detectHistogramStatsDecoding ... Restarting the depth-first walk on each leg of a binary expression is convoluted. ISTM the correct logic is to walk the path backwards to the first relevant function. Signed-off-by: Bryan Boreham <bjboreham@gmail.com>	2025-06-06 14:52:59 +01:00