Outlook and Zimbra Connector not sending attachments

[fultonj]

On which partition did you install zimbra? I don't see a separate partition above and I'd wager you installed it under the / partition which seems to have plenty of inodes, thus I don't don't think you have the inode problem that I did.

[Jrtbloke]

there is only one partition on disk0s2, so i guess i'm adrift again :-( many thanks for your assistance fultonj, i'll try and raise this with support.

[BenG]

Ok, we had this issue with 5.0.20, upgraded to 6.0.4 and 6.0.4 patched ZCO, and it is still there.
Same symptoms as original poster.
We actually set the Timeouts from 60000 down to 30000 and it started working again.... which was really counter intuitive. (We mistyped 300 000).
The only other clue was that a few other things were going on with the DSL link (it is a hosted server) so the link was probably unreliable at the same time.

So 30000 may just be a coincidence.
But the biggest pain with this issue is there is no simple way to determine which email is at fault and/or even if it is happening.
Sometimes the user has no idea until we notice a continuous bandwidth spike and investigate....
So even if we can't fix the sync issue, can we at least fix the error reporting?

Rgds Ben

[fsiegel]

BenG -- I'm going to need a little more info here. You say you have the same symptoms as the original poster. The original post was 2 years ago.

So can you enable ZCO logging (just logging and http logging), try a run (as small a window as possible), and post the logs? If I see that, perhaps I can see what is going on, figure out which email to report, etc.

Thanks

[jpca]

We are having a similar issue... CONSTANT RESYNCHING that fails each time.

ZIMBRA 5.0.15 with Connector for Outlook... also updated to connector for 5.0.21 to see if that would resolve the issue. We've rebooted the affected OUTLOOK mail client computer and we've restarted ZIMBRA

During the synch attempt bandwidth is consuming the entire local up speed while it is sending data to the zimbra server. We pay for BW at 95th%, so this compounds our IP cost as well.

This issue arose when a 6mb file failed to send... continuously retrying... even now after the file was deleted from the OUTLOOK outbox, the server is continuously/repeatedly sending data at full bandwidth and failing each time.

In the meantime, smaller messages, even messages WITH small attachments will send successfully. Another message with a 4mb attachment also fails.

This seems similar to this issue: Constant resynching?

Please advise as to how we can RESOLVE this ASAP.

Thanks much in advance for assistance.

[fsiegel]

We had a bug, 40396, but this was resolved in 5.0.20. If you're having the issue in 5.0.21, I would want to see a ZCO log file in this case. If this message is in your Sent Items folder, delete it and try again.

Also, what is the value of zimbraMtaMaxMessageSize?

[jpca]

zimbraFileUploadMaxSize: 31211520
zimbraMtaMaxMessageSize: 30720000

Size limits look ok.

To clarify:

We are using server 5.0.15... I installed a more recent connector to see if that would resolve the issue.

Are you indicating that this issue is resolved in server 5.0.20? If so, we'll expedite an upgrade to the 5.0.21.

Thanks.

[fsiegel]

Well, we did fix something that looks like this in 5.0.20. But you said you tried 5.0.21, and it didn't work, so I'm not sure. Also, it's not that great an idea to have a newer ZCO than server -- that's not really supported. If you upgrade to 5.0.21, you should upgrade both ZCO and ZCS. In many cases, this will work ok -- but it's not recommended.

But I'm confused. You did try 5.0.21, and you got the same issue?

[jpca]

I figured mixing the server/connector versions was not the best plan... but it does seem to be working in general.

Now since my last post, I did NOT find the outgoing message in the sent folder, however, I DID find it in the trash (I had trashed it).

After emptying the trash, It looks like the synchronize has now completed successfully... AND my it's not trying to resynch again... AND my bandwidth is running at normal levels. So, it was stuck on that message regardless of trying to send or trying to synch the trash folder.

If it helps, the attachment in question was an MP3 file.

SO... for the moment... problem resolved. I'll let you know if there is further issue with large files sending.

Thanks much for the clue that led to resolving this.

[fconil]

Ok, we had this issue with 5.0.20, upgraded to 6.0.4 and 6.0.4 patched ZCO, and it is still there.
Same symptoms as original poster.
We actually set the Timeouts from 60000 down to 30000 and it started working again.... which was really counter intuitive. (We mistyped 300 000).
The only other clue was that a few other things were going on with the DSL link (it is a hosted server) so the link was probably unreliable at the same time.

So 30000 may just be a coincidence.
But the biggest pain with this issue is there is no simple way to determine which email is at fault and/or even if it is happening.
Sometimes the user has no idea until we notice a continuous bandwidth spike and investigate....
So even if we can't fix the sync issue, can we at least fix the error reporting?

Rgds Ben

More on this issue, as I was the one witnessing it first hand:

The user rang me telling me that he could not synchronise his mails.
He was trying to send a 4MB file and it would follow this cycle:
Go all the way to 100% (fairly fast),
hung there for what seemed to be one minute (didn't think of clocking it back then)
Then fails with a status message in the bottom left corner of the zimbra progress window saying "network connection failure".

Checking on the server yielded only one "fileUploaded" error yet sync attempts every 5 minutes or so (but no message regarding files being sent).

I deleted the outbox message, deleted the trash contents, deleted the drafts contents, re-sync'ed with success.

I tried to send a similar email by attaching a compressed version of the afore mentioned file, but the same problem occurred again.

Then after changing the timeout to 300 seconds, yet every other setting to 30sec, I was suddenly able to send 2MB,4MB,9MB files with no problems.

I still wish there was a better way to pinpoint the problem as I can't reproduce it at will.