Details
-
Improvement
-
Resolution: Fixed
-
Critical
-
4.0.0
-
Security Level: Public
Description
I think it would be of great value if we logged operations that are considered as "extremely slow" (takes more than half a second). These numbers can currently be queried from the get timings API in memcached unless the node is restarted. Logging them would make sure support can see them in the data sent to them (and we may correlate it to events in ns_servers log).
Attachments
For Gerrit Dashboard: MB-15368 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
52205,5 | MB-15368: Log operations that take longer than 1s | sherlock | memcached | Status: MERGED | +2 | +1 |
52277,8 | MB-15368: Log operations that take longer than 1s | 3.0 | memcached | Status: MERGED | +2 | +1 |
52424,3 | Merge remote-tracking branch 'couchbase/sherlock' | master | memcached | Status: MERGED | +2 | +1 |