[ceph] Improve health metrics #2852
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
We do it for all the other keys of the raw dict, so let's do it for this one too (even though I'm not a huge fan of this "try/except everything" approach).
Behavior before this commit: the `ceph.num_near_full_osds` and `ceph.num_full_osds` would either: * take the value `0` and be tagged with no `osd` tag if no osd is reporting health issues * take a value representing the usage percentage, tagged by `osd` This doesn't really make sense with respect to the name of the metrics (`num_*`). To solve this, replace these metrics with the following: * `ceph.num_near_full_ods` and `ceph.num_full_osds` report the total number of osds that are respectively near full and full. Not tagged by osd. * when some osds report health issues, the check sends a `ceph.osd.pct_used` metric which reports the usage percentage, tagged by osd. Unfortunately we can't send `0` values on this metric when no osd reports health issues since we can't tag by osd in that case. This should make these metrics more usable. Also, use `gauge` since there's no reason to use `count`.
|
cc @vagelim: Let me know what you think of these changes |
|
This is a much better solution! I was unsure how to send this data to DD in a useful way. Please ping me when these changes have made it to prod, I will need to update the default screenboard accordingly. |
masci
approved these changes
Sep 16, 2016
|
Merging, thanks for the review! @vagelim: this will be released with |
|
The metadata: https://github.com/DataDog/dogweb/pull/15018 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
KeyErroronosd_pool_statsceph.num_full_osdsandceph.num_near_full_osdsmetrics report values that make more sense and addceph.osd.pct_usedmetric (see explanation below).Motivation
Testing
Updated the existing mock tests, and made them more detailed
Additional Notes
health metrics
Previous behavior: the
ceph.num_near_full_osdsandceph.num_full_osdswould either:0and be tagged with noosdtag if no osd isreporting health issues
osdThis doesn't really make sense with respect to the name of the metrics
(
num_*). To solve this, replace these metrics with the following:ceph.num_near_full_odsandceph.num_full_osdsreport the totalnumber of osds that are respectively near full and full. Not tagged
by osd.
ceph.osd.pct_usedmetric which reports the usagepercentage, tagged by osd. Unfortunately we can't send
0valueson this metric when no osd reports health issues since we can't tag by
osd in that case.
This should make these metrics more usable. Also, use
gaugesincethere's no reason to use
count.