[SERVER-20786] Performance Regression in Mongo-Perf MMS/single threaded workloads from 3.0.6 (Update.MmsIncShallow1, ...) Created: 24/Sep/15 Updated: 06/Dec/22 Resolved: 05/Aug/19 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Performance |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | David Daly | Assignee: | Backlog - Query Team (Inactive) |
| Resolution: | Won't Fix | Votes: | 0 |
| Labels: | mpreg, query-44-grooming | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Assigned Teams: |
Query
|
| Operating System: | ALL |
| Sprint: | Performance B (11/02/15) |
| Participants: |
| Description |
|
Performance regression across all 4 build variants for single threaded performance on Mongo-Perf. Slightly more pronounced on 1 node repl set over stand-alone, but present in both cases. Update.MmsIncShallow1 is the largest offender. Regression in WiredTiger Standalone (cumulative 11-13% degradation) and in MMAP (cumulative 13-17% degradation). Can be reproduced with
|
| Comments |
| Comment by Craig Homa [ 05/Aug/19 ] | |||||||||||||||||||||||||||||||||||||||||
|
Closing as future perf work should be based on a broad view of system perf instead of targeting old reports of perf regressions. | |||||||||||||||||||||||||||||||||||||||||
| Comment by Eddie Louie [ 06/Sep/17 ] | |||||||||||||||||||||||||||||||||||||||||
|
This seems like a very old ticket. Should we close this? | |||||||||||||||||||||||||||||||||||||||||
| Comment by David Daly [ 30/Nov/15 ] | |||||||||||||||||||||||||||||||||||||||||
|
This has gotten a couple percent faster (8% regression) since this ticket was filed, but still worth investigating. Here are more recent WT results | |||||||||||||||||||||||||||||||||||||||||
| Comment by Githook User [ 27/Oct/15 ] | |||||||||||||||||||||||||||||||||||||||||
|
Author: {u'username': u'ksuarz', u'name': u'Kyle Suarez', u'email': u'ksuarz@gmail.com'}Message: Closes #1040 Signed-off-by: dalyd <david.daly@mongodb.com> | |||||||||||||||||||||||||||||||||||||||||
| Comment by Githook User [ 22/Oct/15 ] | |||||||||||||||||||||||||||||||||||||||||
|
Author: {u'username': u'dalyd', u'name': u'David', u'email': u'david.daly@mongodb.com'}Message: Merge pull request #1032 from ksuarz/master
| |||||||||||||||||||||||||||||||||||||||||
| Comment by Githook User [ 22/Oct/15 ] | |||||||||||||||||||||||||||||||||||||||||
|
Author: {u'username': u'ksuarz', u'name': u'Kyle Suarez', u'email': u'ksuarz@gmail.com'}Message: | |||||||||||||||||||||||||||||||||||||||||
| Comment by Martin Bligh [ 08/Oct/15 ] | |||||||||||||||||||||||||||||||||||||||||
|
OK, I backed out some logging changes I had in my dev view - now I get 27500 ops/s with either 3.0.6 or master. | |||||||||||||||||||||||||||||||||||||||||
| Comment by Martin Bligh [ 08/Oct/15 ] | |||||||||||||||||||||||||||||||||||||||||
|
Oops. Previous seems to be bleed-through from something else, not sure how perf attributed it to mongod. Old profile:
Sometimes (on new version) I get __log_wrlsn_server traffic, sometimes I don't
Sometimes I get this too (FTDC)
| |||||||||||||||||||||||||||||||||||||||||
| Comment by David Daly [ 06/Oct/15 ] | |||||||||||||||||||||||||||||||||||||||||
|
I did some profiling comparing between 3.1.0 and a recent master. The largest differences I was able to pull out were malloc related.
|