Timeline gives an error after adjusting milestones

[zamoose@…]

I cannot view the timeline now, after having created, renamed and deleted some milestones. I receive the following error now when I try to hit the page:

 Trac detected an internal error:

column rev is not unique

The full trace is shown below:

'''Python traceback'''

Traceback (most recent call last):
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/web/cgi_frontend.py", line 130, in run
    dispatch_request(req.path_info, req, env)
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/web/main.py", line 139, in dispatch_request
    dispatcher.dispatch(req)
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/web/main.py", line 107, in dispatch
    resp = chosen_handler.process_request(req)
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/Timeline.py", line 140, in process_request
    filters)
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/versioncontrol/web_ui/changeset.py", line 116, in get_timeline_events
    chgset = repos.get_changeset(rev)
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/versioncontrol/cache.py", line 42, in get_changeset
    self.sync()
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/versioncontrol/cache.py", line 84, in sync
    "VALUES (%s,%s,%s,%s)", (str(current_rev),
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/db.py", line 219, in execute
    args or [])
  File "/home/zamoose/packages/lib/python2.3/site-packages/trac/db.py", line 211, in _rollback_on_error
    return function(self, *args, **kwargs)
IntegrityError: column rev is not unique

You may view it yourself at ​http://trac.zamoose.org/timeline

[zamoose]

Interesting. It no longer gives the error after editing several Wiki pages. Hmmm. Any idea what went wrong so that I can avoid it in the future?

[cboos]

This had nothing to do with milestones, that was simply a resync attempt triggered by the timeline, which failed because a concurrent resync was already in progress.

The resync should silently complete in this case.

[sid]

Looks pretty similar to #4043

[cboos]

Replying to sid:

Looks pretty similar to #4043

Not a duplicate of that one, because: "It no longer gives the error after editing several Wiki pages".
So it's rather the expected occasional behavior on clashing concurrent resync attempts, as discussed in #4586.

#4043 is certainly related, but on PostgreSQL, there's apparently sometimes a very critical condition as the db remains in an inconsistent state, for some reason, so that every further connection attempt will fail.

[jp [spam] jhcore.com]

I was getting the same error, except I could not navigate anywhere away from it. I had been doing some major subversion directory moving before hand. I fixed it by resync'ing:

trac-admin {site} resync

[cboos]

Replying to jp [spam jhcore.com]:

I was getting the same error, except I could not navigate anywhere away from it. I had been doing some major subversion directory moving before hand. I fixed it by resync'ing:

trac-admin {site} resync

Yes, this is actually the #4043 problem, which was actually not postgresql specific.

There's a fix in preparation and we'll soon release 0.10.4 with that fix.

[sid]

#6227 was marked as duplicate of this ticket.