For: #14355
This commit updates Prometheus to adopt stdlib's log/slog package in
favor of go-kit/log. As part of converting to use slog, several other
related changes are required to get prometheus working, including:
- removed unused logging util func `RateLimit()`
- forward ported the util/logging/Deduper logging by implementing a small custom slog.Handler that does the deduping before chaining log calls to the underlying real slog.Logger
- move some of the json file logging functionality to use prom/common package functionality
- refactored some of the new json file logging for scraping
- changes to promql.QueryLogger interface to swap out logging methods for relevant slog sugar wrappers
- updated lots of tests that used/replicated custom logging functionality, attempting to keep the logical goal of the tests consistent after the transition
- added a healthy amount of `if logger == nil { $makeLogger }` type conditional checks amongst various functions where none were provided -- old code that used the go-kit/log.Logger interface had several places where there were nil references when trying to use functions like `With()` to add keyvals on the new *slog.Logger type
Signed-off-by: TJ Hoplock <t.hoplock@gmail.com>
`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>
* 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>
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>
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>
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>
* TSDB: Document what needs locking in memSeries
* TSDB: Lock around access to series labels
So we can modify them to reset the symbol-table.
* TSDB: Make label locking conditional on build tag
---------
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
Instead of carrying around extra fields in `Meta` structs which let us
approximate what was in the chunk at the time, take a copy of the chunk.
This simplifies lots of code, and lets us correct a couple of tests which
were embedding the wrong answer.
We can also remove boundedIterator, which was only used to constrain
the OOO head chunk.
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
The only call we have to LabelValuesFor() has an index.Postings, and we
expand it to pass to this method, which will iterate over the values.
That's a waste of resources: we can iterate on the index.Postings
directly.
If there's any downstream implementation that has a slice of series,
they can always do an index.ListPostings from them: doing that is
cheaper than expanding an abstract index.Postings.
Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
It's quite common during the compaction cycle to hold series IDs for
series that aren't in the TSDB head anymore.
We shouldn't fail if that happens, as the caller has no way to figure
out which one of the IDs doesn't exist.
Fixes https://github.com/prometheus/prometheus/issues/14278
Signed-off-by: Oleg Zaytsev <mail@olegzaytsev.com>
Add method `PostingsForLabelMatching` to `tsdb.IndexReader`, to obtain postings for labels with a certain name and values accepted by a provided callback, and use it from `tsdb.PostingsForMatchers`.
The intention is to optimize regexp matcher paths, especially not having to load all label values before matching on them.
Plus tests, and refactor some `tsdb/index.Reader` methods.
Benchmarking shows memory reduction up to ~100%, and speedup of up to ~50%.
Signed-off-by: Arve Knudsen <arve.knudsen@gmail.com>
Co-authored-by: Bartlomiej Plotka <bwplotka@gmail.com>
This PR is a reference implementation of the proposal described in #10420.
In addition to what described in #10420, in this PR I've introduced labels.StableHash(). The idea is to offer an hashing function which doesn't change over time, and that's used by query sharding in order to get a stable behaviour over time. The implementation of labels.StableHash() is the hashing function used by Prometheus before stringlabels, and what's used by Grafana Mimir for query sharding (because built before stringlabels was a thing).
Follow up work
As mentioned in #10420, if this PR is accepted I'm also open to upload another foundamental piece used by Grafana Mimir query sharding to accelerate the query execution: an optional, configurable and fast in-memory cache for the series hashes.
Signed-off-by: Marco Pracucci <marco@pracucci.com>
The ChunkReader interface's Chunk() has been changed to ChunkOrIterable().
This is a precursor to OOO native histogram support - with OOO native histograms, the chunks.Meta passed to Chunk() can result in multiple chunks being returned rather than just a single chunk (e.g. if oooMergedChunk has a counter reset in the middle).
To support this, ChunkOrIterable() requires either a single chunk or an iterable to be returned. If an iterable is returned, the caller has the responsibility of converting the samples from the iterable into possibly multiple chunks. The OOOHeadChunkReader now returns an iterable rather than a chunk to prepare for the native histograms case. Also as a beneficial side effect, oooMergedChunk and boundedChunk has been simplified as they only need to implement the Iterable interface now, not the full Chunk interface.
---------
Signed-off-by: Fiona Liao <fiona.y.liao@gmail.com>
Co-authored-by: George Krajcsovits <krajorama@users.noreply.github.com>
Currently memSeries holds a single head chunk in-memory and a slice of mmapped chunks.
When append() is called on memSeries it might decide that a new headChunk is needed to use for given append() call.
If that happens it will first mmap existing head chunk and only after that happens it will create a new empty headChunk and continue appending
our sample to it.
Since appending samples uses write lock on memSeries no other read or write can happen until any append is completed.
When we have an append() that must create a new head chunk the whole memSeries is blocked until mmapping of existing head chunk finishes.
Mmapping itself uses a lock as it needs to be serialised, which means that the more chunks to mmap we have the longer each chunk might wait
for it to be mmapped.
If there's enough chunks that require mmapping some memSeries will be locked for long enough that it will start affecting
queries and scrapes.
Queries might timeout, since by default they have a 2 minute timeout set.
Scrapes will be blocked inside append() call, which means there will be a gap between samples. This will first affect range queries
or calls using rate() and such, since the time range requested in the query might have too few samples to calculate anything.
To avoid this we need to remove mmapping from append path, since mmapping is blocking.
But this means that when we cut a new head chunk we need to keep the old one around, so we can mmap it later.
This change makes memSeries.headChunk a linked list, memSeries.headChunk still points to the 'open' head chunk that receives new samples,
while older, yet to be mmapped, chunks are linked to it.
Mmapping is done on a schedule by iterating all memSeries one by one. Thanks to this we control when mmapping is done, since we trigger
it manually, which reduces the risk that it will have to compete for mmap locks with other chunks.
Signed-off-by: Łukasz Mierzwa <l.mierzwa@gmail.com>
Use AppendableGauge to detect corrupt chunk with gauge histograms.
Detect if first sample is a gauge but the chunk is not set up to contain
gauge histograms.
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
Signed-off-by: George Krajcsovits <krajorama@users.noreply.github.com>
Wiser coders than myself have come to the conclusion that a `switch`
statement is almost always superior to a statement that includes any
`else if`.
The exceptions that I have found in our codebase are just these two:
* The `if else` is followed by an additional statement before the next
condition (separated by a `;`).
* The whole thing is within a `for` loop and `break` statements are
used. In this case, using `switch` would require tagging the `for`
loop, which probably tips the balance.
Why are `switch` statements more readable?
For one, fewer curly braces. But more importantly, the conditions all
have the same alignment, so the whole thing follows the natural flow
of going down a list of conditions. With `else if`, in contrast, all
conditions but the first are "hidden" behind `} else if `, harder to
spot and (for no good reason) presented differently from the first
condition.
I'm sure the aforemention wise coders can list even more reasons.
In any case, I like it so much that I have found myself recommending
it in code reviews. I would like to make it a habit in our code base,
without making it a hard requirement that we would test on the CI. But
for that, there has to be a role model, so this commit eliminates all
`if else` occurrences, unless it is autogenerated code or fits one of
the exceptions above.
Signed-off-by: beorn7 <beorn@grafana.com>
`safeChunk` is only obtained from the `headChunkReader.Chunk` call where
the chunk is already fetched and stored with the `safeChunk`. So, when
getting the iterator for the `safeChunk`, we don't need to get the chunk again.
Also removed a couple of unnecessary fields from `safeChunk` as a part of this.
Signed-off-by: Ganesh Vernekar <ganeshvern@gmail.com>
Instead of passing in a `ScratchBuilder` and `Labels`, just pass the
builder and the caller can extract labels from it. In many cases the
caller didn't use the Labels value anyway.
Now in `Labels.ScratchBuilder` we need a slightly different API: one
to assign what will be the result, instead of overwriting some other
`Labels`. This is safer and easier to reason about.
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
This necessitates a change to the `tsdb.IndexReader` interface:
`index.Reader` is used from multiple goroutines concurrently, so we
can't have state in it.
We do retain a `ScratchBuilder` in `blockBaseSeriesSet` which is
iterator-like.
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
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>
