2.4.6 reconstruct: timestamp mismatch which doesn't go away

Simon Matter simon.matter at invoca.ch
Wed Jan 19 05:00:00 EST 2011

>> Hi,
>> I've done some upgrades from 2.3.16 to 2.4.6 and it seems sometimes a
>> reconstruct is the only way to get a mailbox work after upgrade. While
>> running reconstruct two things have shown up which I have not seen wirh
>> older releases:
>> 1) on a mailbox which was used by a client reconstruct was just sitting
>> there doing nothing. Seems like it was waiting for a lock to be freed?
>> After terminating the client reconstruct would immediately resume it's
>> work. Maybe that's expected, I'm just wondering how those with large
>> number of users will reconstruct while the server is online? I mean,
>> with older releases it was always possible to do a full reconstruct
>> while
>> users had access to their mailboxes. Is it not possible that way
>> anymore?
>> 2) I got some "timestamp mismatch" errors while reconstructing my own
>> mailbox. I've read here
>> http://www.mail-archive.com/cyrus-devel@lists.andrew.cmu.edu/msg01481.html
>> that this should happen only once because the timestamp of the file
>> should
>> be corrected. That doesn't seem to be true in my case. Any ideas why?
> OK, removing cyrus.index and doing the reconstruct makes the error go
> away. But, I was thinking that the files mtime is set to what is found in
> it's "Date:" field? That's however not the case as it seems. I'm really
> not sure I understand the logic of which timestamp is stored where?
> I see that utime() is called on the files while running reconstruct,
> setting the already wrong mtime again?

Replying to myself again :)

Found it, sort of. The messages in question don't have any Received:
headers. They are in the Sent folder and old clients/servers didn't seem
to send Received: headers in that case.
Now, does it make sense to fall back to Date: header in that case to
determine internaldate?

Something like this in message_create_record()?

--- cyrus-imapd-2.4.6/imap/message.c.orig       2010-12-20
14:15:49.000000000 +0100
+++ cyrus-imapd-2.4.6/imap/message.c    2011-01-19 10:51:57.000000000 +0100
@@ -570,6 +570,9 @@
        if (body->received_date)
            newdate = message_parse_date(body->received_date,
+       else if (body->date)
+           newdate = message_parse_date(body->date,
        if (newdate)
            record->internaldate = newdate;

This should apply only to locally copied messages (via IMAP) and therefore
should be quite safe, is it?

Any comments?


More information about the Info-cyrus mailing list