* Extend promtool to support compaction analysis
This commit extends the promtool tsdb analyze command to help
troubleshoot high Prometheus disk usage. The command now plots a
distribution of how full chunks are relative to the maximum capacity of
120 samples per chunk.
Signed-off-by: fpetkovski <filip.petkovsky@gmail.com>
* Update cmd/promtool/tsdb.go
Co-authored-by: Bartlomiej Plotka <bwplotka@gmail.com>
Co-authored-by: Bartlomiej Plotka <bwplotka@gmail.com>
* Fix: Use json.Unmarshal() instead of json.Decoder
See https://ahmet.im/blog/golang-json-decoder-pitfalls/
json.Decoder is for JSON streams, not single JSON objects / bodies.
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Revert modifications to targetgroup parsing
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Add sorting and filtering to flags page
Signed-off-by: Dustin Hooten <dustinhooten@gmail.com>
* Make filter understand
Signed-off-by: Dustin Hooten <dustinhooten@gmail.com>
* split big state object into smaller ones
Signed-off-by: Dustin Hooten <dustinhooten@gmail.com>
* use fuzzy match and sanitize html for search results
Signed-off-by: Dustin Hooten <dustinhooten@gmail.com>
* use fuzzy.filter
Signed-off-by: Dustin Hooten <dustinhooten@gmail.com>
* replace fuzzy lib by @nexucis/fuzzy + fix flags issues
Signed-off-by: Augustin Husson <husson.augustin@gmail.com>
* replace fuzzy by @nexucis/fuzzy in ExpressionInput.tsx
Signed-off-by: Augustin Husson <husson.augustin@gmail.com>
* remove fuzzy lib from package.json
Signed-off-by: Augustin Husson <husson.augustin@gmail.com>
* fix flags test
Signed-off-by: Augustin Husson <husson.augustin@gmail.com>
* simplify the input in the fuzzy search
Signed-off-by: Augustin Husson <husson.augustin@gmail.com>
* cleanup html to be easily compatible with the dark theme
Signed-off-by: Augustin Husson <husson.augustin@gmail.com>
* fix filtering when there is no result
Signed-off-by: Augustin Husson <husson.augustin@gmail.com>
* use id to fix the test
Signed-off-by: Augustin Husson <husson.augustin@gmail.com>
Co-authored-by: Dustin Hooten <dustinhooten@gmail.com>
* Added feature flag support to unit tests
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Added/fixed tests
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Addressed review comments
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Added selection flot plugin
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Added time selection
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Added tests
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Removed irrelevant line in license header
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* promtool: backfill: allow configuring block duration
When backfilling large amounts of data across long periods of time, it
may in certain circumstances be useful to use a longer block duration to
increase the efficiency and speed of the backfilling process. This patch
adds a flag --block-duration-power to allow a user to choose the power N
where the block duration is 2^(N+1)h.
Signed-off-by: Steve Kuznetsov <skuznets@redhat.com>
* promtool: use sub-tests in backfill testing
Signed-off-by: Steve Kuznetsov <skuznets@redhat.com>
* backfill: add messages to tests for clarity
When someone new breaks a test, seeing "expected: false, got: true" is
really not useful. A nice message helps here.
Signed-off-by: Steve Kuznetsov <skuznets@redhat.com>
* backfill: test long block durations
A test that uses a long block duration to write bigger blocks is added.
The check to make sure all blocks are the default duration is removed.
Signed-off-by: Steve Kuznetsov <skuznets@redhat.com>
All this is doing is wrapping the inner alert details display with a
conditional `{open && ...}`.
This already improves https://github.com/prometheus/prometheus/issues/8548 a
lot for cases where there are many individual firing/pending alert elements
under each alerting rule.
E.g. for a list of 200 rules with ~100 alert elements each, this changed the page
render time from 30 seconds to 1s.
Signed-off-by: Julius Volz <julius.volz@gmail.com>
Push updates to the repo sync PRs if there is already a PR open. This
allows for cumulative updates to be synced.
Signed-off-by: SuperQ <superq@gmail.com>
* Added MaxSamplesPerSend
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Added tests
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Fixed order of require
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Added docs
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* writes -> writesReceived
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Improved send loop
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Cut v2.28.0-rc.0 (#8954)
* Cut v2.28.0-rc.0
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Changelog fixup
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Address review comments
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Downgrade some features to enhancements
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Adjust release date to today
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Migrate HTTP SD docs from docs repo (#8972)
See discussion in https://github.com/prometheus/docs/pull/1975
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Cut Prometheus v2.28.0 (#8973)
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Cut v2.28.0-rc.0
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Changelog fixup
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Address review comments
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Downgrade some features to enhancements
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Adjust release date to today
Signed-off-by: Julius Volz <julius.volz@gmail.com>
* Show progress bar when max is not 0
Signed-off-by: Levi Harrison <git@leviharrison.dev>
* Added semicolon
Signed-off-by: Levi Harrison <git@leviharrison.dev>
Add cleanup of the lockfile when the db is cleanly closed
The metric describes the status of the lockfile on startup
0: Already existed
1: Did not exist
-1: Disabled
Therefore, if the min value over time of this metric is 0, that means that executions have exited uncleanly
We can then use that metric to have a much lower threshold on the crashlooping alert:
If the metric exists and it has been zero, two restarts is enough to trigger the alarm
If it does not exist (old prom version for example), the current five restarts threshold remains
Signed-off-by: Julien Duchesne <julien.duchesne@grafana.com>
* Change metric name + set unset value to -1
Signed-off-by: Julien Duchesne <julien.duchesne@grafana.com>
* Only check the last value of the clean start alert
Signed-off-by: Julien Duchesne <julien.duchesne@grafana.com>
* Fix test + nit
Signed-off-by: Julien Duchesne <julien.duchesne@grafana.com>
