Jacopo authored 5 years ago and Peter Leitzen committed 5 years ago

Those feature flags were always enabled so we can remove them safely.

Marcel Amirault authored 5 years ago and Achilleas Pipinellis committed 5 years ago

Square and Angle bracket use causes kramdown warnings,
so changed or enclosed in backticks as needed

David Wilkins authored 5 years ago and Achilleas Pipinellis committed 5 years ago

- Strikeout for metrics that aren't reported anymore

- added  for all the metrics that appear to be still
  available

- Added  for the new metrics.

- Determine Release when the new metrics appeared

- Add descriptions for the new metrics

- remove the adornments  and 

- removce the lines with strikeout

We don't use phase restarts, as we use `preload_app`:
https://github.com/puma/puma/blob/master/README.md#clustered-mode
`:puma_phase` values will always be zero.

Ben Kochie authored 5 years ago and Bob Van Landuyt :neckbeard: committed 5 years ago

* Remove `controller` and `action` labels from duration histogram.
* Create a new simple counter for `controller` and `action`.
* Adjust histogram buckets to observe smaller response times.

Marcel Amirault authored 5 years ago and Achilleas Pipinellis committed 5 years ago

Previously, we used brackets to denote the tier badges,
but this made Kramdown, the docs site Markdown renderer,
show many warnings when building the site. This is now
fixed by using parentheses instead of square brackets.

This was caused by [PREMIUM] looking like a link to
Kramdown, which couldn't find a URL there.

See:
- https://gitlab.com/gitlab-com/gitlab-docs/merge_requests/484
- https://gitlab.com/gitlab-org/gitlab-ce/issues/63800

Ryan Cobb authored 5 years ago and Douwe Maan committed 5 years ago

Previously we were recording process start time as seconds from boot.
This makes it so we record as epoch time.

Marcel Amirault authored 5 years ago and Evan Read committed 5 years ago

Add backticks around square brackets, or fix links, as necessary.

PumaWorkerKiller is used for periodically checking and killing
workers (the biggest one) if overall memory reaches specified
limit. This metric allows us to watch number of killed workers.

Jan Provaznik authored 5 years ago and Mayra Cabrera committed 5 years ago

It assures that requests are aborted after 60 seconds, otherwise
an exception is raised. This exception is logged by Sentry, also
there is a Prometheus counter for measuring number of requests in each
state.

This sampler gathers Puma-specific metrics which can be used by
Prometheus then.

This adds ruby and unicorn instrumentation. This was originally
intended in 11.11 but due to performance concerns it was reverted. This
new commit foregoes the sys-proctable gem was causing performance issues
previously.

This reverts merge request !27474

This adds back ruby_memory_bytes for backwards compatibility, limits
code duplication, cleans up unused methods, and limits the unicorn
worker sampling scope.

This updates monitor docs to reflect the new ruby and unicorn metrics as
well as making it so we fetch process start time via the proc table
instead of via CLOCK_BOOTTIME

Signed-off-by: Takuya Noguchi <takninnovationresearch@gmail.com>

Update the list of GitLab metrics to include the missing metric names.

Now the files are identical again compared to EE.

These are backported from
https://gitlab.com/gitlab-org/gitlab-ee/merge_requests/5050

Was introduced in the time that GitLab still used NFS, which is not
required anymore in most cases. By removing this, the API it calls will
return empty responses. This interface has to be removed in the next
major release, expected to be 12.0.

* Use a simple counter for sampler duration instead of a histogram.
* Use a counter to collect GC time.
* Remove unused objects metric.
* Cleanup metric names to match Prometheus conventions.
* Prefix generic GC stats with `gc_stat`.
* Include worker label on memory and file descriptor metrics.

 - Change the wording a little bit

Ben Kochie authored 7 years ago and Ben Kochie committed 7 years ago

Document the original release of each Prometheus metric.

* Use `_total` naming convention.
* Add to metrics documentation.

Add `seconds` unit to latency metrics based on uptream naming
convention[0].

[0]: https://prometheus.io/docs/practices/naming/#metric-names