On Sat, Aug 13, 2016, at 03:12 AM, Kenneth Marshall wrote: > Hi Ellie, > > I tried the patch and it did print the warnings on the replica. > Unfortunately, > it looked like it was looping on the master (sync_client) and the updates > that > could be made were not. I had to rollback to the dying version. > > Regards, > Ken I've managed to reproduce this (more or less) and I can see what's going on now -- with my patch, if a mailbox on the replica is too old (or contains messages without guids), it won't be able to replicate that whole user. Or at least, it won't be able to replicate that user /at a user level/ -- though their other mailboxes might be replicable if targeted individually with sync_client -m (but I haven't tested this). It fails like this because the sync_server in 2.5 treats any mailbox error as a failure of the entire "GET USER" command. This is different on master: there, mailboxes that have errors during "GET USER" are just omitted from the response, and thus replication proceeds successfully with at least some of the affected user's mailboxes (but will probably error out later in the process, when it tries to upload the "missing" mailboxes). The patch works as I expected under master's semantics, but fails inappropriately under 2.5's semantics. Investigation continues... Cheers, ellie ---- Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
