Commit graph

42 commits

Author SHA1 Message Date
beorn7 865d16f870 Rename Gorilla into varbit 2016-03-23 16:30:41 +01:00
beorn7 c72979e3ed Remove a redundancy from Gorilla-style chunks
So far, the last sample in a chunk was saved twice. That's required
for adding more samples as we need to know the last sample added to
add more samples without iterating through the whole chunk. However,
once the last sample was added to the chunk before it's full, there is
no need to save it twice. Thus, the very last sample added to a chunk
can _only_ be saved in the header fields for the last sample. The
chunk has to be identifiable as closed, then. This information has
been added to the flags byte.
2016-03-20 23:09:48 +01:00
beorn7 9d8fbbe822 Review improvements 2016-03-17 17:31:56 +01:00
beorn7 8cdced3850 Implement Gorilla-inspired chunk encoding
This is not a verbatim implementation of the Gorilla encoding.  First
of all, it could not, even if we wanted, because Prometheus has a
different chunking model (constant size, not constant time).  Second,
this adds a number of changes that improve the encoding in general or
at least for the specific use case of Prometheus (and are partially
only possible in the context of Prometheus). See comments in the code
for details.
2016-03-17 14:47:08 +01:00
beorn7 e7ac9c6863 Improvments based on review
- Moved returns into the default section of switch statement that can
  only happen then.

- Fix typo.
2016-03-17 14:37:24 +01:00
beorn7 161eada3ad Make chunkIterator even leaner. 2016-03-09 16:20:39 +01:00
beorn7 c13b1ecfe9 Make chunk iterators more DRY
This finally extracts all the common code of the two chunk iterators
into one. Any future chunk encodings with fast access by index can use
the same iterator by simply providing an indexAccessor. Other future
chunk encodings without fast index access (like Gorilla-style) can
still implement the chunkIterator interface as usual.
2016-03-07 20:23:14 +01:00
beorn7 32f280a3cd Slim down the chunkIterator interface
For one, remove unneeded methods.

Then, instead of using a channel for all values, use a
bufio.Scanner-like interface. This removes the need for creating a
goroutine and avoids the (unnecessary) locking performed by channel
sending and receiving.

This will make it much easier to write new chunk implementations (like
Gorilla-style encoding).
2016-03-07 19:50:13 +01:00
beorn7 0ea5801e47 Handle errors caused by data corruption more gracefully
This requires all the panic calls upon unexpected data to be converted
into errors returned. This pollute the function signatures quite
lot. Well, this is Go...

The ideas behind this are the following:

- panic only if it's a programming error. Data corruptions happen, and
  they are not programming errors.

- If we detect a data corruption, we "quarantine" the series,
  essentially removing it from the database and putting its data into
  a separate directory for forensics.

- Failure during writing to a series file is not considered corruption
  automatically. It will call setDirty, though, so that a
  crashrecovery upon the next restart will commence and check for
  that.

- Series quarantining and setDirty calls are logged and counted in
  metrics, but are hidden from the user of the interfaces in
  interface.go, whith the notable exception of Append(). The reasoning
  is that we treat corruption by removing the corrupted series, i.e. a
  query for it will return no results on its next call anyway, so
  return no results right now. In the case of Append(), we want to
  tell the user that no data has been appended, though.

Minor side effects:

- Now consistently using filepath.* instead of path.*.

- Introduced structured logging where I touched it. This makes things
  less consistent, but a complete change to structured logging would
  be out of scope for this PR.
2016-03-02 23:02:34 +01:00
beorn7 ce58fd357b Merge branch 'beorn7/storage' into beorn7/storage2
Conflicts:
	storage/local/chunk.go
	storage/local/interface.go
2016-03-02 16:09:32 +01:00
beorn7 2581648f70 Separate iterators by offset
Add test that exposes the problem.
2016-03-02 16:01:03 +01:00
beorn7 b876f8e6a5 Move lastSamplePair method up to memorySeries
This implies a slight change of behavior as only samples added to the
respective instance of a memorySeries are returned. However, this is
most likely anyway what we want.

Following cases:

- Server has been restarted: Given the time it takes to cleanly
  shutdown and start up a server, the series are now stale anyway. An
  improved staleness handling (still to be implemented) will be based
  on tracking if a given target is continuing to expose samples for a
  given time series. In that case, we need a full scrape cycle to
  decide about staleness. So again, it makes sense to consider
  everything stale directly after a server restart.

- Series unarchived due to a read request: The series is definitely
  stale so we don't want to return anything anyway.

- Freshly created time series or series unarchived because of a sample
  append: That happens because appending a sample is imminent. Before
  the fingerprint lock is released, the series will have received a
  sample, and lastSamplePair will always returned the expected value.
2016-02-19 18:16:41 +01:00
beorn7 1e13f89039 Return SamplePair istead of *SamplePair consistently
Formalize ZeroSamplePair as return value for non-existing samples.

Change LastSamplePairForFingerprint to return a SamplePair (and not a
pointer to it), which saves allocations in a potentially extremely
frequent call.
2016-02-19 17:00:40 +01:00
beorn7 d290340367 Fix and improve chunkDesc locking 2016-02-19 16:24:38 +01:00
beorn7 0e202dacb4 Streamline series iterator creation
This will fix issue #1035 and will also help to make issue #1264 less
bad.

The fundamental problem in the current code:

In the preload phase, we quite accurately determine which chunks will
be used for the query being executed. However, in the subsequent step
of creating series iterators, the created iterators are referencing
_all_ in-memory chunks in their series, even the un-pinned ones. In
iterator creation, we copy a pointer to each in-memory chunk of a
series into the iterator. While this creates a certain amount of
allocation churn, the worst thing about it is that copying the chunk
pointer out of the chunkDesc requires a mutex acquisition. (Remember
that the iterator will also reference un-pinned chunks, so we need to
acquire the mutex to protect against concurrent eviction.) The worst
case happens if a series doesn't even contain any relevant samples for
the query time range. We notice that during preloading but then we
will still create a series iterator for it. But even for series that
do contain relevant samples, the overhead is quite bad for instant
queries that retrieve a single sample from each series, but still go
through all the effort of series iterator creation. All of that is
particularly bad if a series has many in-memory chunks.

This commit addresses the problem from two sides:

First, it merges preloading and iterator creation into one step,
i.e. the preload call returns an iterator for exactly the preloaded
chunks.

Second, the required mutex acquisition in chunkDesc has been greatly
reduced. That was enabled by a side effect of the first step, which is
that the iterator is only referencing pinned chunks, so there is no
risk of concurrent eviction anymore, and chunks can be accessed
without mutex acquisition.

To simplify the code changes for the above, the long-planned change of
ValueAtTime to ValueAtOrBefore time was performed at the same
time. (It should have been done first, but it kind of accidentally
happened while I was in the middle of writing the series iterator
changes. Sorry for that.) So far, we actively filtered the up to two
values that were returned by ValueAtTime, i.e. we invested work to
retrieve up to two values, and then we invested more work to throw one
of them away.

The SeriesIterator.BoundaryValues method can be removed once #1401 is
fixed. But I really didn't want to load even more changes into this
PR.

Benchmarks:

The BenchmarkFuzz.* benchmarks run 83% faster (i.e. about six times
faster) and allocate 95% fewer bytes. The reason for that is that the
benchmark reads one sample after another from the time series and
creates a new series iterator for each sample read.

To find out how much these improvements matter in practice, I have
mirrored a beefy Prometheus server at SoundCloud that suffers from
both issues #1035 and #1264. To reach steady state that would be
comparable, the server needs to run for 15d. So far, it has run for
1d. The test server currently has only half as many memory time series
and 60% of the memory chunks the main server has. The 90th percentile
rule evaluation cycle time is ~11s on the main server and only ~3s on
the test server. However, these numbers might get much closer over
time.

In addition to performance improvements, this commit removes about 150
LOC.
2016-02-19 16:24:38 +01:00
beorn7 ef3ab96111 Populate first and last time in the chunk descriptor earlier
The First time is kind of trivial as we always know it when we create
a new chunkDesc.

The last time is only know when the chunk is closed, so we have to set
it at that time.

The change saves a lot of digging down into the chunk
itself. Especially the last time is relative expensive as it involves
the creation of an iterator. The first time access now doesn't require
locking, which is also a nice gain.
2016-02-15 14:06:09 +01:00
Julius Volz 9b6d69610a Fix various typos in comments.
Helpfully reported by
https://goreportcard.com/report/github.com/prometheus/prometheus :)
2016-02-10 03:47:00 +01:00
beorn7 582af1618c Streamline chunk writing
This helps to avoid allocations in the same way we were already doing
it during reading.
2016-01-25 16:36:36 +01:00
Fabian Reinartz 1535ef1457 Replace metric.SamplePair with model.SamplePair 2015-08-22 14:52:35 +02:00
Fabian Reinartz 306e8468a0 Switch from client_golang/model to common/model 2015-08-21 13:33:38 +02:00
beorn7 699946bf32 Fix chunk desc loading.
If all samples in consecutive chunks have the same timestamp, the way
we used to load chunks will fail. With this change, the persist
watermark is used to load the right amount of chunkDescs from disk.

This bug is a possible reason for the rare storage corruption we have
observed.
2015-07-16 13:09:20 +02:00
Fabian Reinartz 6bfb4549a6 storage: add LastSamplePairForFingerprint method 2015-06-23 13:45:15 +02:00
Fabian Reinartz b105e26f4d storage: remove global flags 2015-06-15 19:01:06 +02:00
beorn7 3b9c421a69 Weed out all the [Gg]et* method names.
The only exception is getNumChunksToPersist to avoid naming the struct
member numChunksToPersist in a weird way.
2015-05-20 19:13:06 +02:00
beorn7 cd5574bf8a Make chunk and series iterators more efficient. 2015-05-20 16:19:34 +02:00
beorn7 b02d900e61 Improve chunk and chunkDesc loading.
Also, clean up some things in the code (especially introduction of the
chunkLenWithHeader constant to avoid the same expression all over the place).

Benchmark results:

BEFORE
BenchmarkLoadChunksSequentially     5000            283580 ns/op          152143 B/op        312 allocs/op
BenchmarkLoadChunksRandomly        20000             82936 ns/op           39310 B/op         99 allocs/op
BenchmarkLoadChunkDescs            10000            110833 ns/op           15092 B/op        345 allocs/op

AFTER
BenchmarkLoadChunksSequentially    10000            146785 ns/op          152285 B/op        315 allocs/op
BenchmarkLoadChunksRandomly        20000             67598 ns/op           39438 B/op        103 allocs/op
BenchmarkLoadChunkDescs            20000             99631 ns/op           12636 B/op        192 allocs/op

Note that everything is obviously loaded from the page cache (as the
benchmark runs thousands of times with very small series files). In a
real-world scenario, I expect a larger impact, as the disk operations
will more often actually hit the disk. To load ~50 sequential chunks,
this reduces the iops from 100 seeks and 100 reads to 1 seek and 1
read.
2015-04-13 21:06:04 +02:00
beorn7 5bea942d8e Improve various things around chunk encoding.
A number of mostly minor things:

- Rename chunk type -> chunk encoding.

- After all, do not carry around the chunk encoding to all parts of
  the system, but just have one place where the encoding for new
  chunks is set based on the flag. The new approach has caveats as
  well, but the polution of so many method signatures is worse.

- Use the default chunk encoding for new chunks of existing
  series. (Previously, only new _series_ would get chunks with the
  default encoding.)

- Use an enum for chunk encoding. (But keep the version number for the
  flag, for reasons discussed previously.)

- Add encoding() to the chunk interface (so that a chunk knows its own
  encoding - no need to have that in a different top-level function).

- Got rid of newFollowUpChunk (which would keep the existing encoding
  for all chunks of a time series). Now only use newChunk(), which
  will create a chunk encoding according to the flag.

- Simplified transcodeAndAdd.

- Reordered methods of deltaEncodedChunk and doubleDeltaEncoded chunk
  to match the order in the chunk interface.

- Only transcode if the chunk is not yet half full. If more than half
  full, add a new chunk instead.
2015-03-14 19:03:20 +01:00
beorn7 13fcf1ddbc Implement double-delta encoded chunks. 2015-03-05 20:33:26 +01:00
Bjoern Rabenstein 5859b74f1b Clean up license issues.
- Move CONTRIBUTORS.md to the more common AUTHORS.
- Added the required NOTICE file.
- Changed "Prometheus Team" to "The Prometheus Authors".
- Reverted the erroneous changes to the Apache License.
2015-01-21 20:07:45 +01:00
Bjoern Rabenstein fee88a7a77 Remove the remaining races, new and old.
Also, resolve a few other TODOs.

Change-Id: Icb39b5a5e8ca22ebcb48771cd8951c5d9e112691
2014-12-03 18:07:23 +01:00
Bjoern Rabenstein 7d11019aa2 Squash a few trivial TODOs.
- Delete unneeded file view_adapter.go.
- Assessed that we still need the fingerprints in nodes
  (to create iterators).
- Turned numMemChunkDescs into a metric.

Change-Id: I29be963c795a075ec00c095f76bf26405535609d
2014-11-27 18:26:06 +01:00
Bjoern Rabenstein 14bda4180c Changes after pair code review.
Change-Id: Ib72d40f8e9027818cfbbd32a7a7201eebda07455
2014-11-25 17:12:59 +01:00
Bjoern Rabenstein bb42cc2e2d Evict based on memory pressure. Evict recently used chunks last.
Change-Id: Ie6168f0cdb3917bdc63b6fe15585dd70c1e42afe
2014-11-25 17:10:39 +01:00
Bjoern Rabenstein a617269b12 Avoid unnecessary cloning of the head chunk.
Change-Id: I5da774515d5493166a197b5814d0a720628cfaff
2014-11-25 17:09:04 +01:00
Bjoern Rabenstein 74c9b34a5e Improve storage instrumentation even more.
Add gauge for chunks and chunkdescs in memory (backed by a global
variable to be used later not only for instrumentation but also for
memory management).

Refactored instrumentation code once more (instrumentation.go is back :).

Change-Id: Ife39947e22a48cac4982db7369c231947f446e17
2014-11-25 17:09:04 +01:00
Bjoern Rabenstein 443dd33805 Improve instrumentation in storage.
Also, fix some other minor bugs.

Change-Id: If72f1c058b0f47d3e378fdf80228d7e9a8db06c7
2014-11-25 17:09:04 +01:00
Bjoern Rabenstein 096fa0f8b2 Squash a number of TODOs.
- Staleness delta is no a proper function parameter and not replicated
  from package ast.

- Named type 'chunks' replaced by explicit '[]chunk' to avoid confusion.

- For the same reason, replaced 'chunkDescs' by '[]*chunkDescs'.

- Verified that math.Modf is not a speed enhancement over conversion
  (actually 5x slower).

- Renamed firstTimeField, lastTimeField into chunkFirstTime and
  chunkLastTime.

- Verified unpin() is sufficiently goroutine-safe.

- Decided not to update archivedFingerprintToTimeRange upon series
  truncation and added a rationale why.

Change-Id: I863b8d785e5ad9f71eb63e229845eacf1bed8534
2014-11-25 17:09:04 +01:00
Bjoern Rabenstein 71206dbc06 More code cleanups.
Add license text everywhere.
And others....

Change-Id: I11ccde267a2ef7eb366c4788ba7aeae14ba7545c
2014-11-25 17:07:44 +01:00
Bjoern Rabenstein f5f9f3514a Major code cleanup.
- Make it go-vet and golint clean.
- Add comments, TODOs, etc.

Change-Id: If1392d96f3d5b4cdde597b10c8dff1769fcfabe2
2014-11-25 17:02:53 +01:00
Julius Volz 8dfaa5ecd2 Remove use of freelists for chunk bufs.
Change-Id: Ib887fdb61e1d96da0cd32545817b925ba88831c1
2014-11-25 17:02:00 +01:00
Bjoern Rabenstein d742edfe0d Fix precision loss.
Large delta values often imply a difference between a large base value
and the large delta value, potentially resulting in small numbers with
a huge precision error. Since large delta values need 8 bytes anyway,
we are not even saving memory.

As a solution, always save the absoluto value rather than a delta once
8 bytes would be needed for the delta. Timestamps are then saved as 8
byte integers, while values are always saved as float64 in that case.

Change-Id: I01100d600515e16df58ce508b50982ffd762cc49
2014-11-25 17:02:00 +01:00
Julius Volz e7ed39c9a6 Initial experimental snapshot of next-gen storage.
Change-Id: Ifb8709960dbedd1d9f5efd88cdd359ee9fa9d26d
2014-11-25 17:02:00 +01:00