This is a followup to https://github.com/prometheus/prometheus/pull/2011.
This publishes more of the methods and other names of the chunk code and
moves the chunk code to its own package. There's some unavoidable
ugliness: the chunk and chunkDesc metrics are used by both packages, so
I had to move them to the chunk package. That isn't great, but I don't
see how to do it better without a larger redesign of everything. Same
for the evict requests and some other types.
This is based on https://github.com/prometheus/prometheus/pull/1997.
This adds contexts to the relevant Storage methods and already passes
PromQL's new per-query context into the storage's query methods.
The immediate motivation supporting multi-tenancy in Frankenstein, but
this could also be used by Prometheus's normal local storage to support
cancellations and timeouts at some point.
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).
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.
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.
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 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.
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.
"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.
