Fossil SCM
Very slow checkin on large repository
303b8b118281ca4…
· opened 15 years, 11 months ago
- Type
- Incident
- Priority
- —
- Severity
- Important
- Resolution
- Overcome_By_Events
- Subsystem
- —
- Created
- April 14, 2010 9:16 a.m.
A large repo takes a long time to do a 'checkin' (autosync is off, and the machine is a very fast one with lots of RAM):
time fossil ci -m 'ci time test' New_Version: 4434f3a0489357cd396359b85fcab7f531d5aefdreal 2m44.684s user 0m44.612s sys 0m10.444s
The corresponding "push" was fast: real 0m7.062s
The repository is big, but not enormous:
fossil rstats Number of Artifacts: 2160 2151 full text + 9 delta blobs 198473 bytes average, 428701769 bytes totalNumber Of Checkins: 7 Number Of Files: 37264 Number Of Wiki Pages: 1 Number Of Tickets: 0 Duration Of Project: 2 days
anonymous claiming to be Joerg Sonnenberger added on 2010-12-21 16:29:38: Isn't this a side effect of the setting repo-cksum=on? As it is now optional, I think this can be closed.
anonymous added on 2010-12-28 18:30:52: What Joerg was pointing to requires a newer version of Fossil. Updating to the newest version will provide you the said setting.
joerg added on 2011-03-25 01:01:35 UTC: Addressed by setting repo-cksum to off.