Add more galera metrics #10675
Add more galera metrics #10675
Conversation
|
Hallo, this PR is a draft. Let us know when it's ready for review |
Codecov Report
Flags with carried forward coverage won't be shown. Click here to find out more. |
mysql/datadog_checks/mysql/const.py
Outdated
| 'wsrep_replicated_bytes': ('mysql.galera.wsrep_replicated_bytes', GAUGE), | ||
| 'wsrep_received_bytes': ('mysql.galera.wsrep_received_bytes', GAUGE), | ||
| 'wsrep_received': ('mysql.galera.wsrep_received', GAUGE), | ||
| 'wsrep_local_state': ('mysql.galera.wsrep_local_state', GAUGE), | ||
| 'wsrep_local_cert_failures': ('mysql.galera.wsrep_local_cert_failures', GAUGE), |
There was a problem hiding this comment.
I'm wondering if these 5 new metrics shouldn't be defined as MONOTONIC, because they are all counters of "how many XX (events or bytes) occurred since the node started".
There was a problem hiding this comment.
@jmeunier28 : what do you think of this ?
There was a problem hiding this comment.
good question I would say that makes sense for things that are events e.g.: wsrep_local_cert_failures should be a count
I think wsrep_local_state makes sense as a gauge bc its a state that fluctuates between defined numbers. As far as bytes written/received go, I think it also makes sense to just leave it as a GAUGE assuming the last number you receive from the query is a Total, which it looks like it is
There was a problem hiding this comment.
Ok thanks, I updated wsrep_local_cert_failures to MONOTONIC type and let the others as GAUGE
There was a problem hiding this comment.
I think wsrep_local_state makes sense as a gauge bc its a state that fluctuates between defined numbers.
Ha yes, I didn't check precisely enough last time.
But you're totally right, this metric will go up and down, so gauge 👍
As far as bytes written/received go, I think it also makes sense to just leave it as a GAUGE assuming the last number you receive from the query is a Total, which it looks like it is
Yes, mysql/mariadb returns the total traffic sent since the start of the process.
|
@jmeunier28 thanks for your review what is the next step to merge this PR ? |
|
@aymeric-ledizes sorry for late response here, but if you want to address #10675 (comment) then I can get your PR merged Thanks for adding these new metrics! They look super useful! |
aymeric-ledizes
force-pushed
the
mysql-more-metrics
branch
from
4edcabf to
9adb0ce
Compare
|
Thanks for the approve @jmeunier28 how to merge the PR ? Are the labels ok ? |
* add more galera useful metric for mysql * add wresp_received, wresp_local_state and wresp_local_cert_failures * set wsrep_local_cert_failures as MONOTONIC
yzhan289
changed the title
What does this PR do?
This PR add more useful Galera metrics:
wsrep_local_statewsrep_replicated_byteswsrep_received_byteswsrep_receivedwsrep_local_cert_failuresMotivation
wsrep_local_stateto determine the Galera state (Synced,Joined,Donor/Desynced,Joining)wsrep_received,wsrep_replicated_bytesandwsrep_received_bytesto estimate the gcache size we need to know the amount of data transitingwsrep_local_cert_failuresto monitor the "conflicts" that we never want to have (because it refuses commits) on a production clusterAdditional Notes
Review checklist (to be filled by reviewers)
changelog/andintegration/labels attached