The was a bug (due to confusion?) on the local metadata cache that is cached
by metric family not the series metric name. The fix is to NOT use that local cache
at all (it's still needed for current metadata API implementation, added TODO
on how we can get rid of it).
I went ahead and also rename Metric field in metadata structs to MetricFamily to make
clear it's not always __name__.
Signed-off-by: bwplotka <bwplotka@gmail.com>
Fix issues raised by staticcheck
We are not enabling staticcheck explicitly, though, because it has too many false positives.
---------
Signed-off-by: Arve Knudsen <arve.knudsen@gmail.com>
Returning 0 from Append means 'unknown', so the series is never cached.
Return arbitrary numbers instead.
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
Instead of storing discovered labels on every target, recompute them if
required. The `Target` struct now needs to hold some more data required
to recompute them, such as ScrapeConfig.
This moves the load from every Prometheus all of the time, to just when
someone views Service Discovery in the UI.
The way `PopulateLabels` is used changes; you are no longer expected to
call it with a part-populated `labels.Builder`.
The signature of `Target.Labels` changes to take a `labels.Builder`
instead of a `ScratchBuilder`, for consistency with `DiscoveredLabels`.
This will save a lot of work when many targets are filtered out in
relabeling. Combine with `keep_dropped_targets` to avoid ever computing
most labels for dropped targets.
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
We should never modify (or even shallow copy) Config after config.Load;
added comments and modified GetScrapeConfigs to do so. For GetScrapeConfigs
the validation (even repeated) was likely doing writes (because global fields was 0). We GetScrapeConfigs concurrently
in tests and ApplyConfig causing test races. In prod there were
races but likelyt only to replace 0 with 0, so not too severe.
I removed validation since I don't see anyone using our config.Config without Load.
I had to refactor one test that was doing it, all others use yaml config.
Fixes#15538
Previous attempt: https://github.com/prometheus/prometheus/pull/15634
Signed-off-by: bwplotka <bwplotka@gmail.com>
Don't forget to set `metrics` field as otherwise scraping will lead to a
nil panic in case the body size limit is reached.
Signed-off-by: Giedrius Statkevičius <giedrius.statkevicius@vinted.com>
Resolves: #15433
When I converted prometheus to use slog in #14906, I update both the
`QueryLogger` interface, as well as how the log calls to the
`QueryLogger` were built up in `promql.Engine.exec()`. The backing
logger for the `QueryLogger` in the engine is a
`util/logging.JSONFileLogger`, and it's implementation of the `With()`
method updates the logger the logger in place with the new keyvals added
onto the underlying slog.Logger, which means they get inherited onto
everything after. All subsequent calls to `With()`, even in later
queries, would continue to then append on more and more keyvals for the
various params and fields built up in the logger. In turn, this causes
unbounded growth of the logger, leading to increased memory usage, and
in at least one report was the likely cause of an OOM kill. More
information can be found in the issue and the linked slack thread.
This commit does a few things:
- It was referenced in feedback in #14906 that it would've been better
to not change the `QueryLogger` interface if possible, this PR
proposes changes that bring it closer to alignment with the pre-3.0
`QueryLogger` interface contract
- reverts `promql.Engine.exec()`'s usage of the query logger to the
pattern of building up an array of args to pass at once to the end log
call. Avoiding the repetitious calls to `.With()` are what resolve the
issue with the logger growth/memory usage.
- updates the scrape failure logger to use the update `QueryLogger`
methods in the contract.
- updates tests accordingly
- cleans up unused methods
Builds and passes tests successfully. Tested locally and confirmed I
could no longer reproduce the issue/it resolved the issue.
Signed-off-by: TJ Hoplock <t.hoplock@gmail.com>
Need to extend `newTestScrapeMetrics`` to get at the Registry.
`gatherLabels` function could go upstream to `client_golang`.
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
Before this change, logs would show like:
```
{...,"target":"http://localhost:8080/metrics","!BADKEY":"Get ..."}
```
After this change
```
{...,"msg":"Get ...","job_name":...,"target":...}
```
Signed-off-by: Simon Pasquier <spasquie@redhat.com>
Fix some edge cases when OOO is enabled
Signed-off-by: Vanshikav123 <vanshikav928@gmail.com>
Signed-off-by: Vanshika <102902652+Vanshikav123@users.noreply.github.com>
Signed-off-by: Jesus Vazquez <jesusvzpg@gmail.com>
Co-authored-by: Jesus Vazquez <jesusvzpg@gmail.com>
* NHCB: scrape use state field and not booleans
From comment https://github.com/prometheus/prometheus/pull/14978#discussion_r1800898724
Also make compareLabels read only and move storeLabels to the first
processed classic histogram series.
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
* Speed up TestConvertClassicHistogramsToNHCB 3x
Reduce the startup time and timeouts
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
* lint fix
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
---------
Signed-off-by: György Krajcsovits <gyorgy.krajcsovits@grafana.com>
scrape: Remove implicit fallback to the Prometheus text format
Remove implicit fallback to the Prometheus text format in case of invalid/missing Content-Type and fail the scrape instead. Add ability to specify a `fallback_scrape_protocol` in the scrape config.
---------
Signed-off-by: alexgreenbank <alex.greenbank@grafana.com>
Signed-off-by: Alex Greenbank <alex.greenbank@grafana.com>
Co-authored-by: Björn Rabenstein <beorn@grafana.com>
