Commit Graph

1242 Commits (ca3119bd24c451851c15e47642e85b449552d85c)

Author SHA1 Message Date
Ayoub Mrini 105ab2e95a
fix(test): adjust defer invocations (#14996)
Signed-off-by: machine424 <ayoubmrini424@gmail.com>
2024-09-27 17:13:51 +01:00
Oleg Zaytsev ada8a6ef10
Add some more tests for MemPostings_Delete
Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-09-27 10:14:39 +02:00
Arthur Silva Sens d5f65cfce0
Merge pull request #14694 from prometheus/ct-histogram
Histogram CT Zero ingestion
2024-09-26 12:48:46 -03:00
Arthur Silva Sens 95a53ef982
Join tests for appending float and histogram CTs
Signed-off-by: Arthur Silva Sens <arthursens2005@gmail.com>
2024-09-26 11:29:31 -03:00
Arthur Silva Sens 6bd9b1a7cc
Histogram CT Zero ingestion
Signed-off-by: Arthur Silva Sens <arthursens2005@gmail.com>
2024-09-26 11:29:22 -03:00
Oleg Zaytsev 4fd2556baa
Extract processWithBoundedParallelismAndConsistentWorkers
Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-09-26 15:43:19 +02:00
Oleg Zaytsev ccd0308abc
Don't do anything if MemPostings are empty
Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-09-25 15:00:10 +02:00
Oleg Zaytsev 9c417aa710
Fix deadlock with empty MemPostings
Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-09-25 14:08:50 +02:00
Bryan Boreham 5d8f0ef0c2
Merge pull request #14721 from bboreham/exp-grow-postings
[PERF] TSDB: Grow postings by doubling
2024-09-25 10:47:55 +01:00
Oleg Zaytsev e196b977af
Process MemPostings.Delete() with GOMAXPROCS workers
We are still seeing lock contention on MemPostings.mtx, and MemPostings.Delete() is by far the most expensive operation on that mutex.

This adds parallelism to that method, trying to reduce the amount of time we spend with the mutex held.

Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-09-25 10:38:47 +02:00
Bryan Boreham ca673eb749 Merge remote-tracking branch 'origin/release-2.55' into merge-2.55-into-main
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-22 17:49:34 +01:00
Bryan Boreham 31c5760551
Neater string vs byte-slice conversions (#14425)
unsafe.Slice and unsafe.StringData were added in Go 1.20

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-21 12:19:21 +02:00
Bryan Boreham d42232e178
Merge pull request #14932 from bboreham/chunk-xor-combine-writebits
[PERF] TSDB: Chunk encoding: shorten some write sequences
2024-09-20 17:53:54 +01:00
Bryan Boreham 6f0d6038b7 [BUGFIX] TSDB: Only query chunks up to truncation time (#14948)
If the query overlaps the range currently undergoing compaction, we
should only fetch chunks up to that time. Need to store that min time
in `HeadAndOOOIndexReader`.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-20 17:44:04 +01:00
Bryan Boreham 9215252221
[BUGFIX] TSDB: Only query chunks up to truncation time (#14948)
If the query overlaps the range currently undergoing compaction, we
should only fetch chunks up to that time. Need to store that min time
in `HeadAndOOOIndexReader`.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-20 18:40:17 +02:00
Ganesh Vernekar 5ccb069414 Backward compatibility with upcoming index v3
Signed-off-by: Ganesh Vernekar <ganeshvern@gmail.com>
2024-09-19 10:27:52 +01:00
George Krajcsovits 0d22a91267 Merge pull request #14874 from krajorama/fix-panic-in-ooo-query2
BUGFIX: TSDB: panic in chunk querier
2024-09-19 10:03:53 +01:00
Bryan Boreham e8c2d916ec lint
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-18 15:23:46 +01:00
Bryan Boreham 648a668835 [PERF] Chunk encoding: combine timestamp writes
Instead of a 2-bit write followed by a 14-bit write, do two 8-bit
writes, which goes much faster since it avoids looping.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-18 13:19:21 +01:00
Bryan Boreham b9a9689aae [PERF] Chunk encoding: simplify writeByte
Rather than append a zero then set the value at that position, append the value.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-18 13:19:04 +01:00
Bryan Boreham b65f1b6560 TSDB: Improve xor-chunk benchmarks
Benchmarks must do the same work N times.
Run 3 cases, where the values are constant, vary a bit, and vary a lot.

Also aim for 120 samples same as TSDB default.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-18 13:14:49 +01:00
Bryan Boreham bb47f78929
Merge pull request #14505 from marioferh/improve_performance_regex
[CHANGE] regexp . to match \n and optimize performance
2024-09-18 09:54:16 +01:00
Antoine Pultier d90d0976b5
fix(bstream/writeByte): ensure it appends only one byte (#14854)
fix(bstream/writeByte): ensure it appends only one byte

Signed-off-by: Antoine Pultier <antoine.pultier@sintef.no>
2024-09-17 16:28:33 +02:00
machine424 d1b4312f0a fix(wlog/watcher_test.go): make TestRun_AvoidNotifyWhenBehind more resilient
Signed-off-by: machine424 <ayoubmrini424@gmail.com>
2024-09-17 13:11:04 +02:00
Mario Fernandez 5814920601
Fix: optimize .* regexp performance
Shortcut for `.*` matches newlines as well.
Add preamble change ^(?s:
Add test
dotAll flag por al regex
Add and fix regex tests

Signed-off-by: Mario Fernandez <mariofer@redhat.com>
2024-09-17 12:18:31 +02:00
Bryan Boreham d5f4fabd12
Merge pull request #14911 from bboreham/clarify-postings-benchmark
TSDB: Simplify benchmark regexps
2024-09-17 11:52:13 +02:00
Carrie Edwards 14e3c05ce8
tsdb: Add support for ingestion of out-of-order native histogram samples (#14546)
Add support for ingesting OOO native histograms

* Add flag for enabling and disabling OOO native histogram ingestion
* Update OOO querying tests to include native histogram samples
* Add OOO head tests
* Add test for OOO native histogram counter reset headers

Signed-off-by: Carrie Edwards <edwrdscarrie@gmail.com>
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
Co-authored by: Carrie Edwards <edwrdscarrie@gmail.com>
Co-authored by: Jeanette Tan <jeanette.tan@grafana.com>
Co-authored by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
Co-authored by: Fiona Liao <fiona.liao@grafana.com>
2024-09-17 11:19:06 +02:00
Harry John 919dc0cbc6
storage: Update LabelQuerier interface to return sorted label values (#14849)
* Change LabelQuerier.LabelValues() to return sorted values

---------

Signed-off-by: 🌲 Harry 🌊 John 🏔 <johrry@amazon.com>
2024-09-17 08:55:02 +02:00
Bryan Boreham a8133f3e87 TSDB: Simplify benchmark regexps
Several regexps were coded like `"^.*$"`, which is an unnatural
formulation nobody is likely to use. Inside `NewMatcher`, `^` and `$`
are added anyway, which makes the form in the benchmark redundant.

It even printed it out in the expected way.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-09-16 17:48:05 +01:00
George Krajcsovits 5aa3d8260a
TSDB: OOO native histograms: prep for multiple ooo head chunks (#14850)
* tsdb: mmapCurrentOOOHeadChunk prepare for multiple ooo chunks

Currently float samples can only create a single ooo head chunk, but
native histograms can result in multiple due to counter resets, etc.

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>

* tsdb: getOOOSeriesChunks prepare for multiple ooo chunks

Currently float samples can only create a single ooo head chunk, but
native histograms can result in multiple due to counter resets, etc.

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>

---------

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-09-11 23:55:39 +01:00
Nathan Baulch 50cd453c8f
chore: Fix typos (#14868)
* Fix typos

---------

Signed-off-by: Nathan Baulch <nathan.baulch@gmail.com>
2024-09-10 22:32:03 +02:00
Bryan Boreham 16e5e99546
Merge pull request #14767 from bboreham/fix-encoding-comment
[Comment] Correct the comment on Decbuf.UvarintBytes
2024-09-09 12:52:36 +01:00
György Krajcsovits d3f4e7c223 Remove unnecessary conversion
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-09-09 12:51:02 +02:00
György Krajcsovits 60ab1cc5a5 BUGFIX: TSDB: panic in chunk querier
Followup to #14831

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-09-09 12:43:02 +02:00
George Krajcsovits 536d9f9ce9
BUGFIX: TSDB: panic in query during truncation with OOO head (#14831)
Check if headQuerier is nil before trying to use it.

* TestQueryOOOHeadDuringTruncate: unit test to check query during truncate
Regression test for #14822

* Simulate race between query and Compact()

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-09-05 17:17:42 +01:00
Joshua Hesketh f2064c7987
NH: Do not re-use spans between histograms (#14771)
promql, tsdb (histograms): Do not re-use spans between histograms

When multiple points exist with the same native histogram schemas they
share their spans.
This causes a problem when a native histogram (NH) schema is modified (for example, during
a Sum) then the other NH's with the same spans are also modified. As such,
we should create a new Span for each NH. This will ensure NH's interfaces
are safe to use without considering the effect on other histograms.

At the moment this doesn't present itself as a problem because in all
aggregations and functions operating on native histograms they are copied
by the promql query engine first.

Signed-off-by: Joshua Hesketh <josh@nitrotech.org>

---------

Signed-off-by: Joshua Hesketh <josh@nitrotech.org>
2024-09-04 12:07:16 +02:00
George Krajcsovits 282fb1632a
Merge pull request #14772 from krajorama/fix-mockseriesiterator
Fix: chunkenc.MockSeriesIterator
2024-09-03 16:55:26 +02:00
Arthur Silva Sens 442f24e099
chore: Simplify TestHeadAppender_AppendCTZeroSample (#14812)
Signed-off-by: Arthur Silva Sens <arthursens2005@gmail.com>
2024-09-02 21:30:37 +01:00
Arve Knudsen 2cfc7b244a
Merge pull request #14700 from shandongzhejiang/main
Comments: fix some function names
2024-09-02 18:59:28 +02:00
Oleg Zaytsev ce7d830f1f
Bring back BenchmarkLoadRealWLs (#14757)
This was part of #14525 which was reverted.
I still think that having this benchmark committed in to the repo is
useful.

Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-09-02 17:20:10 +01:00
György Krajcsovits a693dd19f2 Fix: chunkenc.MockSeriesIterator
Starts its index from 0 , but users call Next() before first sample
so it needs to start from -1

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-30 16:44:36 +02:00
Bryan Boreham 0a4f130b39 [Comment] Correct the comment on Decbuf.UvarintBytes
The value is valid when returned, but can become invalid later.

Return to previous wording.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-30 09:40:18 +01:00
Callum Styan a77f5007f9
fix bug with metadata for rw2 (#14766)
Signed-off-by: Callum Styan <callumstyan@gmail.com>
2024-08-30 08:14:20 +01:00
Bryan Boreham 1f38ae7bca [TESTS] TSDB: fix up OOO tests for new Series behaviour
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-29 10:59:09 +01:00
Bryan Boreham cde42f30e9 TSDB: streamline reading of overlapping head chunks
`getOOOSeriesChunks` was already finding sets of overlapping chunks; we
store those in a `multiMeta` struct so that `ChunkOrIterable` can
reconstruct an `Iterable` easily and predictably.

We no longer need a `MergeOOO` flag to indicate that this Meta should
be merged with other ones; this is explicit in the `multiMeta` structure.

We also no longer need `chunkMetaAndChunkDiskMapperRef`.

Add `wrapOOOHeadChunk` to defeat `chunkenc.Pool` - chunks are reset
during compaction, but if we wrap them (like `safeHeadChunk` was doing
then this is skipped) .

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-29 10:57:29 +01:00
Bryan Boreham 838e49e7b8 [REFACTOR] TSDB: move chunkFromSeries from headChunkReader to head
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-29 10:51:48 +01:00
Björn Rabenstein 1d6e0071b7
Merge pull request #14751 from riskrole/main
chore: fix some comments
2024-08-28 16:38:39 +02:00
riskrole 406bf775aa chore: fix some comments
Signed-off-by: riskrole <yuhang@before.tech>
2024-08-28 11:26:57 +08:00
Marco Pracucci ef649d5968
Revert " Store `mmMaxTime` in same field as `seriesShard`"
Signed-off-by: Marco Pracucci <marco@pracucci.com>
2024-08-26 08:56:16 +02:00
Bryan Boreham 33adbe47b1 [PERF] TSDB: Grow postings by doubling
Go's built-in append() grows larger slices with factor 1.3, which means we do a lot more allocating and copying for larger postings.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-24 11:16:58 +01:00
György Krajcsovits 183bbc39a2 Make requesting merge with OOO head explicit in chunk.Meta
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-23 15:50:53 +02:00
György Krajcsovits 41c076196e New cases in Test_ChunkQuerier_OOOQuery and Test_Querier_OOOQuery
Case 1: OOO in-memory head chunk overlaps with first mmaped in-order chunk.

Query: |----------------------------------------------------------------|
InO:    |------mmap---------------||---------mem----------------------|
OOO:     |-----mem-----------|

This triggers ChunkOrIterableWithCopy not including OOO head chunks bug.

Similar to #14693 however testing the end of the interval doesn't
trigger the problem because there the in-order head chunk will be
trimmed with a tombstone, causing the code to switch to ChunkOrIterable
which was fixed.
See a36d1a8a92/tsdb/querier.go (L646)
where len(p.bufIter.Intervals) will be non zero, because it includes the
tombstone to trim the result to the query max time.

Thus a new test is added to check the overlap at the beginning of the
interval that has a separate chunk, which does not need trimming.

Note: same test doesn't fail for sample querier in Test_Querier_OOOQuery
as that doesn't use copy, that is copyHeadChunk is false in the if
condition above.

Case 2:

OOO mmaped head chunk overlaps with first mmaped in-order chunk.

Query: |----------------------------------------------------------------|
InO:    |------mmap---------------||---------mem----------------------|
OOO:     |-----mmap-----------|                             |--mem--|

In this case the meta contains the reference of the in-order chunk and
no indication that a merge is needed with the OOO mmaped chunk.

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-23 15:50:47 +02:00
Arve Knudsen b0aba26ed5 tsdb: Fix ValNone typo in comment
Signed-off-by: Arve Knudsen <arve.knudsen@gmail.com>
2024-08-23 08:20:20 +02:00
beorn7 0f760f63dd lint: Revamp our linting rules, mostly around doc comments
Several things done here:

- Set `max-issues-per-linter` to 0 so that we actually see all linter
  warnings and not just 50 per linter. (As we also set
  `max-same-issues` to 0, I assume this was the intention from the
  beginning.)

- Stop using the golangci-lint default excludes (by setting
  `exclude-use-default: false`. Those are too generous and don't match
  our style conventions. (I have re-added some of the excludes
  explicitly in this commit. See below.)

- Re-add the `errcheck` exclusion we have used so far via the
  defaults.

- Exclude the signature requirement `govet` has for `Seek` methods
  because we use non-standard `Seek` methods a lot. (But we keep other
  requirements, while the default excludes completely disabled the
  check for common method segnatures.)

- Exclude warnings about missing doc comments on exported symbols. (We
  used to be pretty adamant about doc comments, but stopped that at
  some point in the past. By now, we have about 500 missing doc
  comments. We may consider reintroducing this check, but that's
  outside of the scope of this commit. The default excludes of
  golangci-lint essentially ignore doc comments completely.)

- By stop using the default excludes, we now get warnings back on
  malformed doc comments. That's the most impactful change in this
  commit. It does not enforce doc comments (again), but _if_ there is
  a doc comment, it has to have the recommended form. (Most of the
  changes in this commit are fixing this form.)

- Improve wording/spelling of some comments in .golangci.yml, and
  remove an outdated comment.

- Leave `package-comments` inactive, but add a TODO asking if we
  should change that.

- Add a new sub-linter `comment-spacings` (and fix corresponding
  comments), which avoids missing spaces after the leading `//`.

Signed-off-by: beorn7 <beorn@grafana.com>
2024-08-22 17:36:11 +02:00
Bryan Boreham 9a74d53935
[BUGFIX] TSDB: Fix query overlapping in-order and ooo head (#14693)
* tsdb: Unit test query overlapping in order and ooo head

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>

* TSDB: Merge overlapping head chunk

The basic idea is that getOOOSeriesChunks can populate Meta.Chunk, but since
it only returns one Meta per overlapping time-slot, that pointer may end up in a
Meta with a head-chunk ID. So we need HeadAndOOOChunkReader.ChunkOrIterable()
to call mergedChunks in that case.

Previously, mergedChunks was checking that meta.Ref was a valid OOO chunk reference,
but it never actually uses that reference; it just finds all chunks overlapping in time.
So we can delete that code.

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

Co-authored-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-21 14:24:20 +01:00
shandongzhejiang b2712ff284 chore: fix some function names
Signed-off-by: shandongzhejiang <shandongzhejiang@icloud.com>
2024-08-21 11:09:37 +08:00
Arve Knudsen 3a78e76282 Upgrade golangci-lint to v1.60.1
Signed-off-by: Arve Knudsen <arve.knudsen@gmail.com>
2024-08-18 12:13:25 +02:00
Bryan Boreham 87dccb1d1b
Merge pull request #14649 from machine424/ftest
fix(tsdb/db_test.go): close the corrupted chunk after creating it to satisfy Windows FS
2024-08-16 11:57:54 +01:00
Arve Knudsen 66388f706a
Merge pull request #14042 from aknuds1/arve/wlog-histograms
tsdb/wlog: Only treat unknown record types as failure
2024-08-16 12:00:49 +02:00
Björn Rabenstein 1daf7cdd62
Merge pull request #14626 from cuiweiyuan/main
chore: fix some function names
2024-08-15 11:46:21 +02:00
cuiweiyuan 1800af54f0 chore: fix some function names
Signed-off-by: cuiweiyuan <cuiweiyuan@aliyun.com>
2024-08-15 13:57:21 +08:00
Arve Knudsen b5d13a1ab5 Merge remote-tracking branch 'prometheus/main' into arve/wlog-histograms
Signed-off-by: Arve Knudsen <arve.knudsen@gmail.com>
2024-08-14 19:04:53 +01:00
Bryan Boreham 512c67ec26 TSDB: Never go over maximum number of OOO chunks
In `mmapCurrentOOOHeadChunk`, check if the number is at the maximum and
drop the data with an error log. This is not expected to happen as the
maximum is over 8 million; that's 8 years of 1 sample every second.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 13:41:59 +01:00
Bryan Boreham 9135da1e4f TSDB: Review feedback
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>

* Re-enable check in `createHeadWithOOOSamples` which wasn't really broken.
* Move code making `Block` into a `Queryable` into test file.
* Make `getSeriesChunks` return a slice (renamed `appendSeriesChunks`).
* Rename `oooMergedChunks` to `mergedChunks`.
* Improve comment on `ChunkOrIterableWithCopy`.
* Name return values from unpackHeadChunkRef.

Co-authored-by: Oleg Zaytsev <mail@olegzaytsev.com>
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 13:41:44 +01:00
Bryan Boreham 7ffd3ca280 TSDB: Cosmetic: move HeadAndOOO implementations where old code was
This makes the diffs easier to follow.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 13:41:13 +01:00
Bryan Boreham e95607b276 TSDB: Lock round access to labels, where necessary
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 13:41:13 +01:00
Bryan Boreham 26b3de0438 TSDB: Remove OOOHeadIndexReader
Use headIndexReader instead.

OOOCompactionHeadIndexReader needs to be expanded slightly, because it previously delegated to OOOHeadIndexReader.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 13:41:13 +01:00
Bryan Boreham a299c7b6d6 TSDB: Remove OOOHeadChunkReader
Use HeadAndOOOChunkReader instead.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 13:41:13 +01:00
Bryan Boreham e7e50a3afd TSDB: Remove code for querying OOO-head only
Just query via `HeadAndOOOQuerier`, which will skip series where no
in-order chunks are in range.

Now we don't need `OOORangeHead`.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 13:41:13 +01:00
Bryan Boreham 0a2ff76881 TSDB tests: Fix up BenchmarkQueries
Was not working even on main.  Some cases still error.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 13:41:04 +01:00
Bryan Boreham f261597944 TSDB: Fix up LabelValues to work for OOO-only head
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 11:19:02 +01:00
Bryan Boreham 6529d6336c TSDB: NewHeadAndOOOChunkReader takes headChunkReader
So we can pass nil and have it read just OOO chunks.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 11:19:02 +01:00
Bryan Boreham e04d137649 [PERF] TSDB: Query head and ooo-head together
Add `HeadAndOOOQuerier` which iterates just once over series, then
where necessary merges chunks from in-order and out-of-order lists.

Add a ChunkQuerier for in-order and ooo together

Add copy-last-chunk behaviour to HeadAndOOOChunkReader

Out-of-order chunk IDs are distinguished from in-order by setting bit 23.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 11:19:02 +01:00
Bryan Boreham da31da3ea6 Refactor: extract selectSeriesSet and selectChunkSeriesSet
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 11:19:02 +01:00
Bryan Boreham 7e24844d08 Refactor: extract headChunkReader.chunkFromSeries()
For when you have a series locked already.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 11:19:02 +01:00
Bryan Boreham a32aca0cd7 Refactoring: extract getOOOSeriesChunks
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 11:19:02 +01:00
Bryan Boreham c75c8f8329 Refactoring: extract getSeriesChunks
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 11:19:02 +01:00
Bryan Boreham 0c852680bf [Benchmark] TSDB: Add BenchmarkQuerierSelectWithOutOfOrder
Refactor existing BenchmarkQuerierSelect to provide the set-up.

Note that Head queries now run faster because they use a RangeHead.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-14 11:19:02 +01:00
György Krajcsovits 41656162fc tsdb: prepare inserting native histograms into OOO head
Rename a variable.
Add parameters to memSeries.insert function.

No effect on how float samples are handled.

Related to #14546

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-14 11:13:47 +02:00
Bryan Boreham aa4b056ad0
Merge pull request #13200 from bboreham/wlog-defer
tsdb/wlog: close segment files sooner
2024-08-13 14:11:38 +01:00
machine424 82f38d3e9a
fix(tsdb/db_test.go): close the corrupted chunk after creating it to satisfy Windows FS
Signed-off-by: machine424 <ayoubmrini424@gmail.com>
2024-08-09 14:53:57 +02:00
George Krajcsovits cf62fb5c44
Merge pull request #14629 from krajorama/fix-to-encoded-chunks
Fix ToEncodedChunks minT for recoded chunks
2024-08-08 20:00:31 +02:00
György Krajcsovits 1ea3781699 Fix ToEncodedChunks minT for recoded chunks
Discovered while working on #14546 OOO native histograms.
Not triggered on main before #14546 as the code path is unused.

There was a bug where the min time of a chunk was adjusted even
if it was only recoded and not completely new.

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-08 15:22:46 +02:00
Ben Ye b7a58dcf3d
Add hidden flag to disable overlapping compaction (#14581)
TSDB: add hidden flag to disable overlapping compaction

Signed-off-by: Ben Ye <benye@amazon.com>

---------

Signed-off-by: Ben Ye <benye@amazon.com>
2024-08-08 12:09:39 +02:00
George Krajcsovits 3a673cd0bc
Merge pull request #14598 from krajorama/fix-compaction-panic
Fix: panic: runtime error: index out of range [4] with length 4
2024-08-07 17:14:14 +02:00
machine424 92873d3009 feat: allow to delay head compaction start time helping Prometheus instances to
avoid simultaneous compactions and reduce stress on shared resources.

This is enabled via `--enable-feature=delayed-compaction`.

Signed-off-by: machine424 <ayoubmrini424@gmail.com>
2024-08-07 17:10:27 +02:00
Oleg Zaytsev 0833d2a230
Fix appendable: check whether last val was a histogram (#14613)
* Fix appendable: check whether last val was a histogram

When appending a float, we were checking whether lastValue was equal to
current value, but we didn't check whether last value was a float value.

Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-08-07 15:02:59 +02:00
György Krajcsovits 98ecdf3589 Fix corrupting spans via iterator sharing
Iterator may share spans without copy, so we always have to make a copy
before modification - copy-on-write.

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-06 16:51:20 +02:00
György Krajcsovits d2f6fa7289 Fix lint error
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-06 13:24:46 +02:00
György Krajcsovits 1b6d1366d8 Fix re-code histogram and chunk re-code conflict
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-06 13:09:17 +02:00
György Krajcsovits aff089a014 Reproduce recoding bug with new and missing buckets
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-06 10:51:44 +02:00
Bryan Boreham 80adc5baf4 Merge remote-tracking branch 'origin/main' into merge-2.54-to-main 2024-08-06 09:19:55 +01:00
machine424 9e43ad2e37 chore(remote_write): clean up as watcher.go is part of wlog now
Signed-off-by: machine424 <ayoubmrini424@gmail.com>
2024-08-05 13:40:23 +02:00
Bryan Boreham 015638c4b6 [BUGFIX] TSDB: Exclude OOO chunks mapped after compaction starts
Otherwise the writer can end up with invalid chunks.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-05 10:35:34 +01:00
Bryan Boreham bded853035 [Test] TSDB: TestOOOCompaction with samples added after compaction starts
Test fails due to bug.

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
2024-08-05 10:35:34 +01:00
George Krajcsovits 00ab05c3b9
Native histograms: fix spurios counter reset when merging recoded chunk to normal chunk (#14513)
* chunkenc: allow missing empty buckets on histogram append

Allow appending to chunks when the histogram to be added is missing
some buckets, but the missing buckets are empty in the chunk.
For example bucket at index 5 is present in the chunk, but its value
is 0 and the new histogram doesn't have a bucket at index 5.

This fixes an issue of merging chunks where one chunk was recoded to
retroactively have some empty buckets in all the histograms and we are
merging in a histogram that doesn't have the empty bucket (because it
was not recoded yet).

The operation alters the histogram that is being added, however this has
already been the case when appending gauge histograms. Thus the test
TestHistogramSeriesToChunks in storage package is changed to explicitly
test what happened to the appended histogram - Compact(0) call is removed.

The new expandIntSpansAndBuckets and expandFloatSpansAndBuckets functions
are a merge of expandSpansForward and counterResetInAnyBucket and
counterResetInAnyFloatBucket.

Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
2024-08-01 09:22:32 +02:00
Bartlomiej Plotka 6816149852
Merge pull request #14525 from colega/merge-mmmaxtime-into-shardhash
Store `mmMaxTime` in same field as `seriesShard`
2024-07-31 08:39:38 +02:00
Max Amin 84b819a69f
feat: add Google cloud roundtripper for remote write (#14346)
* feat: Google Auth for remote write

Signed-off-by: Max Amin <maxamin@google.com>

---------

Signed-off-by: Max Amin <maxamin@google.com>
2024-07-30 16:25:19 +01:00
Oleg Zaytsev 0300ad58a9
Revert the option regardless of error
Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-07-30 11:31:31 +02:00
Oleg Zaytsev d8e1b6bdfd
Store mmMaxTime in same field as seriesShard
We don't use seriesShard during DB initialization, so we can use the
same 8 bytes to store mmMaxTime, and save those during the rest of the
lifetime of the database.

This doesn't affect CPU performance.

Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
2024-07-30 10:20:29 +02:00