PMM-9288 Extended profiler. #3662
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## v3 #3662 +/- ##
==========================================
+ Coverage 43.78% 43.89% +0.10%
==========================================
Files 367 367
Lines 44538 44627 +89
==========================================
+ Hits 19503 19591 +88
Misses 23353 23353
- Partials 1682 1683 +1
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
| bucket.Mongodb.MDocsScannedCnt = float32(v.Count) // TODO: Check is it right value | ||
| bucket.Mongodb.MDocsScannedMax = float32(v.Scanned.Max) | ||
| bucket.Mongodb.MDocsScannedMin = float32(v.Scanned.Min) | ||
| bucket.Mongodb.MDocsScannedP99 = float32(v.Scanned.Pct99) | ||
| bucket.Mongodb.MDocsScannedSum = float32(v.Scanned.Total) |
JiriCtvrtka
requested review from
ademidoff,
idoqo and
igroene
and removed request for
a team
| bucket.Mongodb.MFullScanSum = float32(v.CollScanSum) / 1000 | ||
| bucket.Mongodb.MFullScanSum = float32(v.CollScanCount) // Sum is same like count in this case |
There was a problem hiding this comment.
@igroene please check if in case of Full Scan (CollScan in Mongo) are Sum and Count values same.
There was a problem hiding this comment.
what if we count the distinct number of collscans and use sum for the total number? would it be too complicated?
There was a problem hiding this comment.
I'm not sure if I understood you correctly, but:
- Currently, it is the total sum of the number of collection scans. Is this what you meant?
- If you're asking about the total number of all stages, then yes, it would be complicated at this point. To achieve that, we would need to change the way data is collected, which could be done in a separate ticket.
Or did you mean something entirely different?
There was a problem hiding this comment.
what I mean is, keep the SUM as the total number of recorded COLSCANS for that particular query shape. Use COUNT for the number of recorded COLSCANS of the particular query shape that have different values for the variables. Assuming we captured the following 4 queries:
db.testcol.find({a:101})
db.testcol.find({a:101})
db.testcol.find({a:200})
db.testcol.find({a:140})
assuming all the above are collscans, SUM would be 4 and COUNT would be 3. Makes sense?
There was a problem hiding this comment.
I see. Yes, this should be possible. Let me adjust the code to make it work. Thanks for the explanation.
There was a problem hiding this comment.
@JiriCtvrtka current implementation should make it, no adjustment required
There was a problem hiding this comment.
Thank you for verifying.
PMM-9288
Link to the Feature Build: SUBMODULES-0
Percona Toolkit: PMM-9288 Extended profiler. percona-toolkit#924
Grafana Dashboards: PMM-9288 Extended profiler. grafana-dashboards#1656
FB: PMM-9288 Extended profiler. Percona-Lab/pmm-submodules#3856