production loggerhead branch leaks memory
Bug #156453 reported by
Max Kanat-Alexander
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
loggerhead |
Fix Released
|
Critical
|
Max Kanat-Alexander |
Bug Description
I'm running mwhudson's "production" branch of loggerhead on my server. I have three auto-published folders and one set of non-auto-published folders.
After running for over a week, the process was consuming 1.1GB of RAM and bzr had become so slow on that server that I couldn't do remote checkouts. (Possibly because the server only has 2GB of RAM and it was swapping.)
Changed in loggerhead: | |
importance: | Undecided → Critical |
status: | New → Confirmed |
Changed in loggerhead: | |
status: | Triaged → In Progress |
Changed in loggerhead: | |
assignee: | rockstar → nobody |
status: | In Progress → Triaged |
Changed in loggerhead: | |
status: | Fix Committed → Fix Released |
Changed in loggerhead: | |
status: | Fix Committed → Confirmed |
Changed in loggerhead: | |
assignee: | nobody → Max Kanat-Alexander (mkanat) |
tags: | added: canonical-losa-lp |
To post a comment you must log in.
I've observed this too. Our solution has been to restart loggerhead every so often :/
Roughly how many branches total do you publish (i.e. how many auto-published branches are there)?