Commit graph

726 commits

Author SHA1 Message Date
Bryan Boreham aafef011b7 Promql: reuse LabelBuilder in aggregations
We have a LabelBuilder in EvalNodeHelper; use it instead of creating a new one at every step.

Need to take some care that different uses of enh.lb do not overlap.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-12-23 13:21:29 +00:00
Bryan Boreham 2c382f5e24 promql: extract function to initialize LabelBuilder
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-12-23 13:21:22 +00:00
Bryan Boreham ccea61c7bf
Merge pull request #11717 from bboreham/labels-abstraction
Add and use abstractions over labels.Labels
2022-12-20 17:23:39 +00:00
Bryan Boreham dbd7021cc2
promql: add test for race conditions in query engine (#11743)
* promql: refactor BenchmarkRangeQuery so we can re-use test cases

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* promql: add test for race conditions in query engine

Note we skip large count_values queries -
`count_values` allocates a slice per unique value in the output, and
this test has unique values on every step of every series so it adds up
to a lot of slices. Add Go runtime overhead for checking `-race`, and
it chews up many gigabytes.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* TestConcurrentRangeQueries: wait before starting goroutine

Instead of starting 100 goroutines which just wait for the semaphore.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-12-20 17:58:29 +01:00
Bryan Boreham aa634e0b7e Update package promql tests for new labels.Labels type
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-12-19 15:22:09 +00:00
Bryan Boreham 56fefcd812 Update package promql for new labels.Labels type
We use `labels.Builder` to parse metrics, to avoid depending on the
internal implementation. This is not efficient, but the feature is only
used in tests. It wasn't efficient previously either - calling `Sort()`
after adding each label.

`createLabelsForAbsentFunction` also uses a Builder now, and gets
an extra `map` to replace the previous `Has()` usage.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

Fix up promql to compile with changes to Labels
2022-12-19 15:22:09 +00:00
Bryan Boreham 1695a7ee2f promql: refactor BenchmarkRangeQuery so we can re-use test cases
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-12-15 19:39:46 +00:00
Bryan Boreham 463f5cafdd storage: re-use iterators to save garbage
Re-use previous memory if it is already of the correct type.

In `NewListSeries` we hoist the conversion to an interface value out
so it only allocates once.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-12-15 18:32:45 +00:00
Bryan Boreham 3c7de69059 storage: allow re-use of iterators
Patterned after `Chunk.Iterator()`: pass the old iterator in so it
can be re-used to avoid allocating a new object.

(This commit does not do any re-use; it is just changing all the method
signatures so re-use is possible in later commits.)

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-12-15 18:32:45 +00:00
Alan Protasio 8460807475 fix blank lines
Signed-off-by: Alan Protasio <approtas@amazon.com>
2022-12-14 13:24:10 -08:00
Alan Protasio f8f4ac14a8 Finishing evalSpanTimer always before return
Signed-off-by: Alan Protasio <approtas@amazon.com>
2022-12-14 13:10:35 -08:00
Bryan Boreham 6bdecf377c
Switch from 'sanity' to more inclusive lanuage (#9376)
* Switch from 'sanity' to more inclusive lanuage

"Removing ableist language in code is important; it helps to create and
maintain an environment that welcomes all developers of all backgrounds,
while emphasizing that we as developers select the most articulate,
precise, descriptive language we can rather than relying on metaphors.

The phrase sanity check is ableist, and unnecessarily references mental
health in our code bases. It denotes that people with mental illnesses
are inferior, wrong, or incorrect, and the phrase sanity continues to be
used by employers and other individuals to discriminate against these
people."

From https://gist.github.com/seanmhanson/fe370c2d8bd2b3228680e38899baf5cc

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-11-28 17:09:18 +00:00
Ganesh Vernekar bde500e690
Merge pull request #11420 from jesusvazquez/jvp/update-sparsehistogram-with-main-v2
Merge branch 'main' into sparsehistogram
2022-10-12 11:45:50 +05:30
beorn7 bf0847073d histogram: Modify getBound to deal properly with infinity
The bucket receiving math.MaxFloat64 observations now has
math.MaxFloat64 as upper bound, while the bucket after it (the last
possible bucket) has +Inf.

This also adds a test for getBound and moves the getBound code to
generic.go (where it should have been in the first place).

Signed-off-by: beorn7 <beorn@grafana.com>
2022-10-06 17:40:03 +02:00
Jesus Vazquez e934d0f011 Merge 'main' into sparsehistogram
Signed-off-by: Jesus Vazquez <jesus.vazquez@grafana.com>
2022-10-05 22:14:49 +02:00
Björn Rabenstein dccfb9db4e
histogram: Remove code replication via generics (#11361)
* histogram: Simplify iterators

We don't really need currLower and currUpper and can calculate it when
needed (as already done for the floatBucketIterator). The calculation
is cheap, while keeping those extra variables around costs RAM
(potentially a lot with many iterators).

* histogram: Convert Bucket/FloatBucket to one generic type

* histogram: Move some bucket iterator code into generic base iterator

* histogram: Remove cumulative iterator for FloatHistogram

We added it in the past for completeness (Histogram has one), but it
has never been used. Plus, even the cumulative iterator for Histogram
is only there for test reasons.

We can always add it back, and then maybe even using generics.

Signed-off-by: beorn7 <beorn@grafana.com>
2022-10-03 16:45:27 +05:30
Giedrius Statkevičius a1d6ba59ac
promql: pass down subquery interval (#11163)
If we are populating series for a subquery then set the interval
parameter accordingly so that downstream users could use that
information.

Signed-off-by: Giedrius Statkevičius <giedrius.statkevicius@vinted.com>
2022-09-30 20:13:38 +05:30
Bryan Boreham 3330d85ba8
Replace sort.Strings and sort.Ints with faster slices.Sort (#11318)
Use new experimental package `golang.org/x/exp/slices`.

slices.Sort works on values that are directly comparable, like ints,
so avoids the overhad of an interface call to `.Less()`.

Left tests unchanged, because they don't need the speed and it may be
a cross-check that slices.Sort gives the same answer.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-09-30 20:03:56 +05:30
beorn7 a7c519930e histograms: Add Compact method to the normal integer Histogram
And use the new method to call to compact Histograms during
parsing. This happens for both `Histogram` and `FloatHistogram`. In
this way, if targets decide to optimize the exposition size by merging
spans with empty buckets in between, we still get a normalized
results. It will also normalize away any valid but weird
representations like empty spans, spans with offset zero, and empty
buckets at the start or end of a span.

The implementation seemed easy at first as it just turns the
`compactBuckets` helper into a generic function (which now got its own
file). However, the integer Histograms have delta buckets instead of
absolute buckets, which had to be treated specially in the generic
`compactBuckets` function. To make sure it works, I have added plenty
of explicit tests for `Histogram` in addition to the `FloatHistogram`
tests.

I have also updated the doc comment for the `Compact` method.

Based on the insights now expressed in the doc comment, compacting
with a maxEmptyBuckets > 0 is rarely useful. Therefore, this commit
also sets the value to 0 in the two cases we were using 3 so far. We
might still want to reconsider, so I don't want to remove the
maxEmptyBuckets parameter right now.

Signed-off-by: beorn7 <beorn@grafana.com>
2022-09-27 13:04:16 +02:00
Bryan Boreham b01d29cf9e promql: in tests use labels.FromStrings
And a few cases of `EmptyLabels()`.
Replacing code which assumes the internal structure of `Labels`.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-09-09 13:34:49 +02:00
Julien Pivotto 96d5a32659
Update go to 1.19, set min version to 1.18 (#11279)
* Update go to 1.19, set min version to 1.18

Signed-off-by: Julien Pivotto <roidelapluie@o11y.eu>

* Update golangci-lint

Signed-off-by: Julien Pivotto <roidelapluie@o11y.eu>

Signed-off-by: Julien Pivotto <roidelapluie@o11y.eu>
2022-09-07 11:30:48 +02:00
Ganesh Vernekar 71489d0e3d
Fix count() for histograms and add test case
Signed-off-by: Ganesh Vernekar <ganeshvern@gmail.com>
2022-08-29 19:57:29 +05:30
Bryan Boreham 8b863c42dd
Optimise relabeling by re-using memory (#11147)
* model/relabel: Add benchmark

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* model/relabel: re-use Builder across relabels

Saves memory allocations.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* labels.Builder: allow re-use of result slice

This reduces memory allocations where the caller has a suitable slice available.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* model/relabel: re-use source values slice

To reduce memory allocations.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* Unwind one change causing test failures

Restore original behaviour in PopulateLabels, where we must not overwrite the input set.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* relabel: simplify values optimisation

Use a stack-based array for up to 16 source labels, which will be the
vast majority of cases.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* lint

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-08-19 15:27:52 +05:30
Ganesh Vernekar 9325caa41c
Remove a TODO that is no longer valid (#11186)
Signed-off-by: Ganesh Vernekar <ganeshvern@gmail.com>

Signed-off-by: Ganesh Vernekar <ganeshvern@gmail.com>
2022-08-18 22:47:12 +05:30
Cosrider af79dce8ea
Close file descriptor in query logging (#11148)
Signed-off-by: Cosrider <cosrider7@gmail.com>

Signed-off-by: Cosrider <cosrider7@gmail.com>
2022-08-17 15:27:27 +05:30
beorn7 c9fd3c235d Merge branch 'main' into sparsehistogram 2022-08-10 17:54:37 +02:00
Vilius Pranckaitis 4660656312
Allow setting custom lookback delta for instant queries (#9946)
* Allow setting custom lookback delta for instant queries

Signed-off-by: Vilius Pranckaitis <vpranckaitis@gmail.com>
2022-08-02 11:15:39 +02:00
Levi Harrison 77a7af4461
Add histogram validation (#11052)
* Add histogram validation

Signed-off-by: Levi Harrison <git@leviharrison.dev>

* Correct negative offset validation

Signed-off-by: Levi Harrison <git@leviharrison.dev>

* Address review comments

Signed-off-by: Levi Harrison <git@leviharrison.dev>

* Validation benchmark

Signed-off-by: Levi Harrison <git@leviharrison.dev>

* Add more checks

Signed-off-by: Levi Harrison <git@leviharrison.dev>

* Attempt to fix tests

Signed-off-by: Levi Harrison <git@leviharrison.dev>

* Fix stuff

Signed-off-by: Levi Harrison <git@leviharrison.dev>
2022-07-29 09:52:49 -05:00
Julien Pivotto d41e5a5582
Prettifier: Add spaces with non-callable keywords (#11005)
* Prettifier: Add spaces with non-callable keywords

I prefer to have a difference between, on one side: functions calls, end(), start(), and on the other side with, without, ignoring, by and group_rrigt, group_left.

The reasoning is that the former ones are not calls, while other are
functions. Additionally, it matches the examples in our documentation.

Signed-off-by: Julien Pivotto <roidelapluie@o11y.eu>

* Fix tests

Signed-off-by: Julien Pivotto <roidelapluie@o11y.eu>
2022-07-15 00:09:56 +02:00
Łukasz Mierzwa 54a3c3ba3f
Print query that caused a panic (#10995)
We print the stacktrace of a panic when query causes one, but there's no
information about the query itself, which makes it harder to debug and
reproduce the issue.
This adds the 'expr' string to the logged panic.

Signed-off-by: Łukasz Mierzwa <l.mierzwa@gmail.com>
2022-07-14 15:04:15 +05:30
beorn7 53982c3562 Merge branch 'main' into sparsehistogram 2022-07-13 21:27:20 +02:00
Oleg Zaytsev d2abe9a58a
Shortcut ContainsSameLabelset() (#11000)
* Shortcut Matrix.ContainsSameLabelset()

It's quite often to execute this check on a Matrix that has zero or only
one series. There's no need to allocate a map for those cases.

There's also a one-liner for two-series case, so why not using it?

Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>

* Add license header

Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>

* Optimize Vector.ContainsSameLabelset

Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2022-07-13 15:18:10 +05:30
beorn7 28f028e938 Merge branch 'main' into sparsehistogram 2022-07-12 19:07:13 +02:00
Harkishen Singh 44fcf876ca
Adds support for prettifying PromQL expression (#10544)
* Implement Pretty() function for AST nodes.

Signed-off-by: Harkishen-Singh <harkishensingh@hotmail.com>

This commit adds .Pretty() for all nodes of PromQL AST.
Each .Pretty() prettifies the node it belongs to, and under
no circustance, the parent or child node is touch/prettified.

Read more in the "Approach" part in `prettier.go`

* Refactor functions between printer.go & prettier.go

Signed-off-by: Harkishen-Singh <harkishensingh@hotmail.com>

This commit removes redundancy between printer.go and prettier.go
by taking out the common code into separate private functions.

* Add more unit tests for Prettier.

Signed-off-by: Harkishen-Singh <harkishensingh@hotmail.com>

* Add support for spliting function calls with 1 arg & unary expressions.

Signed-off-by: Harkishen-Singh <harkishensingh@hotmail.com>

This commit does 2 things:
1. It adds support to split function calls that have 1 arg and exceeds the max_characters_per_line
to multiple lines.
2. Splits Unary expressions that exceed the max_characters_per_line. This is done by formatting the child node
and then removing the prefix indent, which is already applied before the unary operator.
2022-07-07 18:13:36 +05:30
beorn7 9eafed0f79 promql: Add histogram_count and histogram_sum
This follow a simple function-based approach to access the count and
sum fields of a native Histogram. It might be more elegant to
implement “accessors” via the dot operator, as considered in the
brainstorming doc [1]. However, that would require the introduction of
a whole new concept in PromQL. For the PoC, we should be fine with the
function-based approch. Even the obvious inefficiencies (rate'ing a
whole histogram twice when we only want to rate each the count and the
sum once) could be optimized behind the scenes.

Note that the function-based approach elegantly solves the problem of
detecting counter resets in the sum of observations in the case of
negative observations. (Since the whole native Histogram is rate'd,
the counter reset is detected for the Histogram as a whole.)

We will decide later if an “accessor” approach is really needed. It
would change the example expression for average duration in
functions.md from

      histogram_sum(rate(http_request_duration_seconds[10m]))
	/
      histogram_count(rate(http_request_duration_seconds[10m]))

to

      rate(http_request_duration_seconds.sum[10m])
	/
      rate(http_request_duration_seconds.count[10m])

[1]: https://docs.google.com/document/d/1ch6ru8GKg03N02jRjYriurt-CZqUVY09evPg6yKTA1s/edit

Signed-off-by: beorn7 <beorn@grafana.com>
2022-06-28 18:16:48 +02:00
beorn7 a3a8f58bb3 promql: Add histogram_fraction function
Signed-off-by: beorn7 <beorn@grafana.com>
2022-06-28 15:58:03 +02:00
beorn7 ffaabea91a promql: Refine zero bucket treatment in histogramQuantile
Essentially, this mirrors the existing behavior for negative buckets:
If a histogram has only negative buckets, the upper bound of the zero
bucket is assumed to be zero.

Furthermore, it makes sure that the zero bucket boundaries are not
modified if a histogram that has no buckets at all but samples in the
zero bucket.

Also, add an TODO to vet if we really want this behavior.

Signed-off-by: beorn7 <beorn@grafana.com>
2022-06-19 15:06:51 +02:00
beorn7 40ad5e284a Merge branch 'main' into beorn7/sparsehistogram 2022-06-09 20:50:30 +02:00
Matthieu MOREL 0906f2eafa
refactor (promql): move from github.com/pkg/errors to 'errors' and 'fmt' (#10817)
Signed-off-by: Matthieu MOREL <mmorel-35@users.noreply.github.com>

Co-authored-by: Matthieu MOREL <mmorel-35@users.noreply.github.com>
2022-06-08 10:47:52 +02:00
Bryan Boreham 2e2c014d52
Labels: optimise creation of signature with/without labels (#10667)
* Labels: create signature with/without labels

Instead of creating a new Labels slice then converting to signature,
go directly to the signature and save time.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* Labels: refactor Builder tests

Have one test with a range of cases, and have them check the final
output rather than checking the internal structure of the Builder.

Also add a couple of cases where the value is "", which should be
interpreted as 'delete'.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* Labels: add 'Keep' function to Builder

This lets us replace `Labels.WithLabels` with the more general `Builder`.

In `engine.resultMetric()` we can call `Keep()` instead of checking
and calling `Del()`.

Avoid calling `Sort()` in `Builder.Labels()` if we didn't add anything,
so that `Keep()` has the same performance as `WithLabels()`.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2022-06-07 10:08:27 +05:30
Łukasz Mierzwa 08262454a3
Preallocate Labels in labels.Builder (#10749)
This tries to avoid re-allocations of labels slice since we know possible max size

Signed-off-by: Łukasz Mierzwa <l.mierzwa@gmail.com>
2022-05-25 16:22:47 +02:00
Ivo Gosemann e22b54e253 Adds day_of_year function to PromQL
Signed-off-by: Ivo Gosemann <ivo.gosemann@sap.com>
2022-05-20 14:08:34 +02:00
beorn7 3bc711e333 Merge branch 'main' into sparsehistogram 2022-05-04 13:37:13 +02:00
beorn7 d16b314b72 Histogram: Do not render empty buckets in JSON output
While empty buckets can make sense in the internal representation (by
joining spans that would otherwise need more overhead for separate
representation), there are no spans in the JSON rendering. Therefore,
the JSON should not contain any empty buckets, since any buckets not
included in the output counts as empty anyway.

This changes both the inefficient MarshalJSON implementation as well
as the jsoniter implementation.

Signed-off-by: beorn7 <beorn@grafana.com>
2022-05-03 18:18:55 +02:00
Matthieu MOREL e2ede285a2
refactor: move from io/ioutil to io and os packages (#10528)
* refactor: move from io/ioutil to io and os packages
* use fs.DirEntry instead of os.FileInfo after os.ReadDir

Signed-off-by: MOREL Matthieu <matthieu.morel@cnp.fr>
2022-04-27 11:24:36 +02:00
beorn7 37bbc07118 Histogram: Add jsoniter marshaling
This now even enables jsoniter marshaling of Points in an instant
query (which previously used the traditional JSON marshaling).

Signed-off-by: beorn7 <beorn@grafana.com>
2022-04-26 15:19:59 +02:00
beorn7 99894f6afa Histogram: Implement inefficient JSON rendering
Format is according to example 4 in
https://docs.google.com/document/d/1Efu0LX-fgNWix6ehfeCR0FzeWtHvftWFNoy7cYW9nqU/edit#

This is inefficient as it doesn't use jsoniter. It actually prevents
it for regular Points (with a conventional float Value), too.

Signed-off-by: beorn7 <beorn@grafana.com>
2022-04-14 20:38:49 +02:00
Alan Protasio ce6a643ee8
Changing TotalQueryableSamples from int to int64 (#10549)
* Changing TotalQueryableSamples from int to int64

Signed-off-by: Alan Protasio <approtas@amazon.com>
2022-04-12 01:22:25 +02:00
beorn7 106e20cde5 Histogram: Fix and simplify histogram_quantile
For conventional histograms, we need to gather all the individual
bucket timeseries at a data point to do the quantile calculation. The
code so far mirrored this behavior for the new native
histograms. However, since a single data point contains all the
buckets alreade, that's actually not needed. This PR simplifies the
code while still detecting a mix of conventional and native
histograms.

The weird signature calculation for the conventional histograms is
getting even weirder because of that. If this PR turns out to do the
right thing, I will implement a proper fix for the signature
calculation upstream.

Signed-off-by: beorn7 <beorn@grafana.com>
2022-04-11 20:53:57 +02:00
beorn7 7ee1836ef5 Merge branch 'main' into sparsehistogram 2022-04-05 18:31:19 +02:00