See discussion in
https://groups.google.com/forum/#!topic/prometheus-developers/bkuGbVlvQ9g
The main idea is that the user of a storage shouldn't have to deal with
fingerprints anymore, and should not need to do an individual preload
call for each metric. The storage interface needs to be made more
high-level to not expose these details.
This also makes it easier to reuse the same storage interface for remote
storages later, as fewer roundtrips are required and the fingerprint
concept doesn't work well across the network.
NOTE: this deliberately gets rid of a small optimization in the old
query Analyzer, where we dedupe instants and ranges for the same series.
This should have a minor impact, as most queries do not have multiple
selectors loading the same series (and at the same offset).
tl;dr: This is not a fundamental solution to the indexing problem
(like tindex is) but it at least avoids utilizing the intersection
problem to the greatest possible amount.
In more detail:
Imagine the following query:
nicely:aggregating:rule{job="foo",env="prod"}
While it uses a nicely aggregating recording rule (which might have a
very low cardinality), Prometheus still intersects the low number of
fingerprints for `{__name__="nicely:aggregating:rule"}` with the many
thousands of fingerprints matching `{job="foo"}` and with the millions
of fingerprints matching `{env="prod"}`. This totally innocuous query
is dead slow if the Prometheus server has a lot of time series with
the `{env="prod"}` label. Ironically, if you make the query more
complicated, it becomes blazingly fast:
nicely:aggregating:rule{job=~"foo",env=~"prod"}
Why so? Because Prometheus only intersects with non-Equal matchers if
there are no Equal matchers. That's good in this case because it
retrieves the few fingerprints for
`{__name__="nicely:aggregating:rule"}` and then starts right ahead to
retrieve the metric for those FPs and checking individually if they
match the other matchers.
This change is generalizing the idea of when to stop intersecting FPs
and go into "retrieve metrics and check them individually against
remaining matchers" mode:
- First, sort all matchers by "expected cardinality". Matchers
matching the empty string are always worst (and never used for
intersections). Equal matchers are in general consider best, but by
using some crude heuristics, we declare some better than others
(instance labels or anything that looks like a recording rule).
- Then go through the matchers until we hit a threshold of remaining
FPs in the intersection. This threshold is higher if we are already
in the non-Equal matcher area as intersection is even more expensive
here.
- Once the threshold has been reached (or we have run out of matchers
that do not match the empty string), start with "retrieve metrics
and check them individually against remaining matchers".
A beefy server at SoundCloud was spending 67% of its CPU time in index
lookups (fingerprintsForLabelPairs), serving mostly a dashboard that
is exclusively built with recording rules. With this change, it spends
only 35% in fingerprintsForLabelPairs. The CPU usage dropped from 26
cores to 18 cores. The median latency for query_range dropped from 14s
to 50ms(!). As expected, higher percentile latency didn't improve that
much because the new approach is _occasionally_ running into the worst
case while the old one was _systematically_ doing so. The 99th
percentile latency is now about as high as the median before (14s)
while it was almost twice as high before (26s).
If the chunks of a series in the checkpoint are all older then the
latest chunk on disk, the head chunk is persisted and therefore has to
be declared closed.
It would be great to have a test for this, but that would require more
plumbing, subject of #447.
PromQL only requires a much narrower interface than local.Storage in
order to run queries. Narrower interfaces are easier to replace and
test, too.
We could also change the web interface to use local.Querier, except that
we'll probably use appending functions from there in the future.
On Windows, it is not possible to rename or delete a file that is
currerntly open. This change closes the file in dropAndPersistChunks
before it tries to delete it, or rename the temporary file to it.
With a lot of series accessed in a short timeframe (by a query, a
large scrape, checkpointing, ...), there is actually quite a
significant amount of lock contention if something similar is running
at the same time.
In those cases, the number of locks needs to be increased.
On the same front, as our fingerprints don't have a lot of entropy, I
introduced some additional shuffling. With the current state, anly
changes in the least singificant bits of a FP would matter.
But only on DEBUG level.
Also, count and report the two cases of out-of-order timestamps on the
one hand and same timestamp but different value on the other hand
separately.
Before, we checkpointed after every newly detected fingerprint
collision, which is not a problem as long as collisions are
rare. However, with a sufficient number of metrics or particular
nature of the data set, there might be a lot of collisions, all to be
detected upon the first set of scrapes, and then the checkpointing
after each detection will take a quite long time (it's O(n²),
essentially).
Since we are rebuilding the fingerprint mapping during crash recovery,
the previous, very conservative approach didn't even buy us
anything. We only ever read from the checkpoint file after a clean
shutdown, so the only time we need to write the checkpoint file is
during a clean shutdown.
Prometheus is Apache 2 licensed, and most source files have the
appropriate copyright license header, but some were missing it without
apparent reason. Correct that by adding it.
The chunk encoding was hardcoded there because it mostly doesn't
matter what encoding is chosen in that test. Since type 1 is
battle-hardened enough, I'm switching to type 2 here so that we can
catch unexpected problems as a byproduct. My expectation is that the
chunk encoding doesn't matter anyway, as said, but then "unexpected
problems" contains the word "unexpected".
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.
This improves fuzz testing in two ways:
(1) More realistic time stamps. So far, the most common case in
practice was very rare in the test: Completely regular increases of
the timestamp.
(2) Verify samples by scanning through the whole relevant section of
the series.
For Gorilla-like chunks, this showed two things:
(1) With more regularly increasing time stamps, BenchmarkFuzz is
essentially as fast as with the traditional chunks:
```
BenchmarkFuzzChunkType0-8 2 972514684 ns/op 83426196 B/op 2500044 allocs/op
BenchmarkFuzzChunkType1-8 2 971478001 ns/op 82874660 B/op 2512364 allocs/op
BenchmarkFuzzChunkType2-8 2 999339453 ns/op 76670636 B/op 2366116 allocs/op
```
(2) There was a bug related to when and how the chunk footer is
overwritten to make use for the last sample. This wasn't exposed by
random access as the last sample of a chunk is retrieved from the
values in the header in that case.
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.
It is now also used in label matching, so the name of the metric
changed from `prometheus_local_storage_invalid_preload_requests_total`
to `non_existent_series_matches_total'.
Only return an error where callers are doing something with it except
simply logging and ignoring.
All the errors touched in this commit flag the storage as dirty
anyway, and that fact is logged anyway. So most of what is being
removed here is just log spam.
As discussed earlier, the class of errors that flags the storage as
dirty signals fundamental corruption, no even bubbling up a one-time
warning to the user (e.g. about incomplete results) isn't helping much
because _anything_ happening in the storage has to be doubted from
that point on (and in fact retroactively into the past, too). Flagging
the storage dirty, and alerting on it (plus marking the state in the
web UI) is the only way I can see right now.
As a byproduct, I cleaned up the setDirty method a bit and improved
the logged errors.
WIP: This needs more tests.
It now gets a from and through value, which it may opportunistically
use to optimize the retrieval. With possible future range indices,
this could be used in a very efficient way. This change merely applies
some easy checks, which should nevertheless solve the use case of
heavy rule evaluations on servers with a lot of series churn.
Idea is the following:
- Only archive series that are at least as old as the headChunkTimeout
(which was already extremely unlikely to happen).
- Then maintain a high watermark for the last archival, i.e. no
archived series has a sample more recent than that watermark.
- Any query that doesn't reach to a time before that watermark doesn't
have to touch the archive index at all. (A production server at
Soundcloud with the aforementioned series churn and heavy rule
evaluations spends 50% of its CPU time in archive index
lookups. Since rule evaluations usually only touch very recent
values, most of those lookup should disappear with this change.)
- Federation with a very broad label matcher will profit from this,
too.
As a byproduct, the un-needed MetricForFingerprint method was removed
from the Storage interface.
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.
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).
I needed this today for debugging. It can certainly be improved, but
it's already quite helpful.
I refactored the reading of heads.db files out of persistence, which
is an improvement, too.
I made minor changes to the cli package to allow outputting via the
io.Writer interface.
Obviously, it's really bad to depend on timing here. The proper fix
would be to have something like WaitForIndexing for other things to
wait for, too.
For now, let's see if the wait time increase fixes the issue.
This fixes https://github.com/prometheus/prometheus/issues/1059 , but
not in the obvious way (simply not updating the persist watermark,
because that's actually not that simple - we don't really know what
has gone wrong exactly). As any errors relevant here are most likely
caused by severe and unrecoverable problems with the series file,
Using the now quarantine feature is the right step. We don't really
have to be worried about any inconsistent state of the series because
it will be removed for good ASAP. Another plus is that we don't have
to declare the whole storage dirty anymore.
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.
Fixes https://github.com/prometheus/prometheus/issues/1401
This remove the last (and in fact bogus) use of BoundaryValues.
Thus, a whole lot of unused (and arguably sub-optimal / ugly) code can
be removed here, too.
In a way, our instants were also ranges, just with the staleness delta
as range length. They are no treated equally, just that in one case,
the range length is set as range, in the other the staleness
delta. However, there are "real" instants where start and and time of
a query is the same. In those cases, we only want to return a single
value (the one closest before or at the equal start and end time). If
that value is the last sample in the series, odds are we have it
already in the series object. In that case, there is no need to pin or
load any chunks. A special singleSampleSeriesIterator is created for
that. This should greatly speed up instant queries as they happen
frequently for rule evaluations.
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.
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.
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.
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.
This gives up on the idea to communicate throuh the Append() call (by
either not returning as it is now or returning an error as
suggested/explored elsewhere). Here I have added a Throttled() call,
which has the advantage that it can be called before a whole _batch_
of Append()'s. Scrapes will happen completely or not at all. Same for
rule group evaluations. That's a highly desired behavior (as discussed
elsewhere). The code is even simpler now as the whole ingestion buffer
could be removed.
Logging of throttled mode has been streamlined and will create at most
one message per minute.
Since we are not overestimating the number of chunks to persist
anymore, this commit also adjusts the default value for
-storage.local.memory-chunks. Update of documentation will follow.
"Rushed mode" is formerly known as "degraded mode", which is changed
with this commit, too. The name "degraded" was very misleading.
Also, switch into rushed mode if we have too many chunks in memory and
an at least reasonable amount of chunks to persist so that speeding up
persisting chunks can help.
If only very few chunks are to be truncated from a very large series
file, the rewrite of the file is a lorge overhead. With this change, a
certain ratio of the file has to be dropped to make it happen. While
only causing disk overhead at about the same ratio (by default 10%),
it will cut down I/O by a lot in above scenario.
Allows to use graphite over tcp or udp. Metrics labels
and values are used to construct a valid Graphite path
in a way that will allow us to eventually read them back
and reconstruct the metrics.
For example, this metric:
model.Metric{
model.MetricNameLabel: "test:metric",
"testlabel": "test:value",
"testlabel2": "test:value",
)
Will become:
test:metric.testlabel=test:value.testlabel2=test:value
escape.go takes care of escaping values to match Graphite
character set, it basically uses percent-encoding as a fallback
wich will work pretty will in the graphite/grafana world.
The remote storage module also has an optional 'prefix' parameter
to prefix all metrics with a path (for example, 'prometheus.').
Graphite URLs are simply in the form tcp://host:port or
udp://host:port.
Because the InfluxDB client library currently pulls in multiple MBs of
unnecessary dependencies, I have modified and cut up the vendored
version to only pull in the few pieces that are actually needed.
On InfluxDB's side, this dependency issue is tracked in:
https://github.com/influxdb/influxdb/issues/3447
Hopefully, it will be resolved soon.
If a password is needed for InfluxDB, it may be supplied via the
INFLUXDB_PW environment variable.
The test had become flaky with Go1.5.
Theory here is that with Go1.5.x, sleeping for 10ms might not be
enough to wake up another goroutine, possibly because it is used for
GC. 50ms should always be enough due to GC pause guarantees with the
new GC.
This is with `golint -min_confidence=0.5`.
I left several lint warnings untouched because they were either
incorrect or I felt it was better not to change them at the moment.
If users see the crash recovery error, the chances are
they aren't shutting down Prometheus correctly. Telling
them how to do so will help them debug and fix the problem.
Perhaps it would be even better to still warn in case the sample value has
changed but the timestamps are equal, but we don't have efficient access
to the last value.
Allow scrape_configs to have an optional proxy_url option which specifies
a proxy to be used for all connections to hosts in that config.
Internally this modifies the various client functions to take a *url.URL pointer
which currently must point to an HTTP proxy (but has been left open-ended to
allow the url format to be extended to support others, such as maybe SOCKS if
needed).
For the label matching index-based preselection phase, don't do an OR
between equality and non-equality matchers. Execute only one of the two
(with equality matchers preferred when present).
Fixes https://github.com/prometheus/prometheus/issues/924
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.
Fixes https://github.com/prometheus/prometheus/issues/481
While doing so, clean up and fix a few other things:
- Fix `go vet` warnings (@fabxc to blame ;).
- Fix a racey problem with unarchiving: Whenever we unarchive a
series, we essentially want to do something with it. However, until
we have done something with it, it appears like a series that is
ready to be archived or even purged. So e.g. it would be ignored
during checkpointing. With this fix, we always load the chunkDescs
upon unarchiving. This is wasteful if we only want to add a new
sample to an archived time series, but the (presumably more common)
case where we access an archived time series in a query doesn't
become more expensive.
- The change above streamlined the getOrCreateSeries ond
newMemorySeries flow. Also, the modTime is now always set correctly.
- Fix the leveldb-backed implementation of KeyValueStore.Delete. It
had the wrong behavior of still returning true, nil if a
non-existing key has been passed in.
See https://github.com/prometheus/prometheus/issues/887, which will at
least be partially fixed by this.
From the spec https://golang.org/ref/spec#Conversions:
"In all non-constant conversions involving floating-point or complex
values, if the result type cannot represent the value the conversion
succeeds but the result value is implementation-dependent."
This ended up setting the converted values to 0 on Debian's Go 1.4.2
compiler, at least on 32-bit Debians.
This commit adds the honor_labels and params arguments to the scrape
config. This allows to specify query parameters used by the scrapers
and handling scraped labels with precedence.
Change #704 introduced a regression that started reading the queue only
after potential crash recovery. When more than the queue capacity was
indexed, Prometheus deadlocked.
This change is conceptually very simple, although the diff is large. It
switches logging from "github.com/golang/glog" to
"github.com/prometheus/log", while not actually changing any log
messages. V(1)-style logging has been changed to be log.Debug*().
This commit creates a (so far unused) package. It contains the a custom
lexer/parser for the query language.
ast.go: New AST that interacts well with the parser.
lex.go: Custom lexer (new).
lex_test.go: Lexer tests (new).
parse.go: Custom parser (new).
parse_test.go: Parser tests (new).
functions.go: Changed function type, dummies for parser testing (barely changed/dummies).
printer.go: Adapted from rules/ and adjusted to new AST (mostly unchanged, few additions).
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.
The one central sample ingestion channel has caused a variety of
trouble. This commit removes it. Targets and rule evaluation call an
Append method directly now. To incorporate multiple storage backends
(like OpenTSDB), storage.Tee forks the Append into two different
appenders.
Note that the tsdb queue manager had its own queue anyway. It was a
queue after a queue... Much queue, so overhead...
Targets have their own little buffer (implemented as a channel) to
avoid stalling during an http scrape. But a new scrape will only be
started once the old one is fully ingested.
The contraption of three pipelined ingesters was removed. A Target is
an ingester itself now. Despite more logic in Target, things should be
less confusing now.
Also, remove lint and vet warnings in ast.go.
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.
