Mailbox Server Timeout and Unable to Login to Webmail After Empty Large Trash

[det]

Hi,

What are the operations involved when emptying a trash? I encountered this issue whereby the mailbox server timeout after a user empty a 3GB trash.

The MTA server sending to the Mailbox server receives a timeout from the mailbox server. Then, users with large mailbox size (>=3GB) on that Mailbox server are unable to login to webmail. The other users with small mailbox size on the same Mailbox server are still able to login to webmail.

Is there any optimization that can be done on the delete operation? Why are users with large mailbox are not able to login to webmail but no issue with IMAP?

Any ideas?

Thanks,
Det

[phoenix]

Surely the question should be, why has the user got 3GB plus in the Trash folder? That folder is (should be) automatically emptied on a regular basis and not contain that much data (the default is it does not contain more than 30 days of data).

[det]

Hi,

Yes it is emptied on a regular basis.

For several days he was testing something and receiving thousands of short emails. He is sending them to trash. It accumulated till it get to 200K emails in the trash.

Then when he empty it, users with large mailbox size (>=3GB) on that Mailbox server are unable to login to webmail but can still IMAP. The other users with small mailbox size on the same Mailbox server are able to login to webmail.

I have to restart zmcontrol to stop the issue and resume operation to normal.

[phoenix]

Deleting that amount of mail will thrash the HD system, that's probably what was causing the problem. The user would have been best to leave the emptying of Trash to Zimbra.

[det]

I see...

Is there anything else that can be done to cut the delete process aside from restarting the zimbra service? Would there be a better option aside from restarting the zimbra service?

[veronica]

Should enable message aging on COS.

[phoenix]

Originally Posted by det

I see...

Is there anything else that can be done to cut the delete process aside from restarting the zimbra service? Would there be a better option aside from restarting the zimbra service?

Assuming you've not modified the message lifetime in the COS (it's set at 30 days on my system) I'd just leave Zimbra to remove them as they age (again, assuming it's not cause a space problem for the server).

[veronica]

By default message aging is disabled.

[det]

Thanks guys. But let's say take aside the trash folder. Let's say it was in one of the folders and he emptied that amount. Wouldn't the issue be the same? Is there a way to get around it?

[veronica]

I dont think the issue is because of emptying folder. Its because of something else. Can you share mailbox.log for that period.