Fix CouchDB 2+ message_queues metric collection #19521
Conversation
At the moment, the `count` value in the metric block is recorded as a gauge. That value is the number of instances of each type of message queue. Say you have two database files open, then `message_queues.couch_file.count` will be `2`. That information is already available in the `_stats` endpoint as `open_os_files`. The actually valuable part of the message_queue metric is the `max` entry, it shows that the longest backlog of messages for each instance type. A longer backlog means a reduction in performance. Without this patch, this value is not available. Technically all of the `max`, `min`, `50`, `90`, `99` are valuable, but I wanted to keep this patch as small as possible until it is clear the change will be accepted. I’m happy to add the other fields as well if desired.
I could find no mention of |
|
The second commit includes the other useful metrics, but do feel free to skip that if the change seems too big |
|
@janl Thanks for the contribution, can I trouble you to take a look at the failing tests and update them? If not I can take a stab at it later this week. |
|
@steveny91 I would, but I need clarification. |
janl
force-pushed
the
fix-couchdb-message-queues
branch
from
0359bf8 to
3cda929
Compare
|
@janl Sorry I missed that question. So essentially our CI relies heavily on a tool called ddev: We then ship tests that are located in these files: Which test functionality of the code. Given that new metrics were added, two tests
https://github.com/DataDog/integrations-core/tree/master/keda/changelog.d. But I can also do so if you would allow me to push to your fork/branch We greatly appreciate the contribution! |
| couchdb.erlang.message_queues.max,gauge,,,,,0,couchdb,, | ||
| couchdb.erlang.message_queues.min,gauge,,,,,0,couchdb,, | ||
| couchdb.erlang.message_queues.50,gauge,,,,,0,couchdb,, | ||
| couchdb.erlang.message_queues.90,gauge,,,,,0,couchdb,, | ||
| couchdb.erlang.message_queues.99,gauge,,,,,0,couchdb,, |
There was a problem hiding this comment.
Non blocking nit: can you also leave a brief description describing these metrics?
At the moment, the
countvalue in the metric block is recorded as a gauge. That value is the number of instances of each type of message queue. Say you have two database files open, thenmessage_queues.couch_file.countwill be2. That information is already available in the_statsendpoint asopen_os_files.The actually valuable part of the message_queue metric is the
maxentry, it shows that the longest backlog of messages for each instance type. A longer backlog means a reduction in performance. Without this patch, this value is not available.Technically all of the
max,min,50,90,99are valuable, but I wanted to keep this patch as small as possible until it is clear the change will be accepted.I’m happy to add the other fields as well if desired.
What does this PR do?
Motivation
Review checklist (to be filled by reviewers)
qa/skip-qalabel if the PR doesn't need to be tested during QA.backport/<branch-name>label to the PR and it will automatically open a backport PR once this one is merged