Postoffice Connector Queue hangs intermittently

Raise/discuss any potential issues with MailEnable for consideration in project issue register.
Post Reply
R4LRetro
Posts: 49
Joined: Wed Mar 25, 2020 1:32 pm

Postoffice Connector Queue hangs intermittently

Post by R4LRetro »

I have emailed support about this issue but wanted to post here in case anyone else was having this issue.

Running Enterprise 10.48 on Server 2022.

Starting yesterday, a user reported that he did not have any emails from the previous night. Literally as he reported the issue, they suddenly appeared in his inbox. Checking our server I noticed the postoffice outbound queue on our server has been piling up, resulting in alerts being sent:

Code: Select all

MEPOCS Outbound Queue Length
MEPOCS outbound queue is experiencing a spike in volume.

MEPOCS Outbound Queue Last Poll
MEPOCS appears to be having difficulty processing the outbound queue.

MEPOCS Outbound Queue Item Max Age
MEPOCS appears to be having difficulty processing the outbound queue.
Some emails in the queue have a status of 'Delivering' but never seem to leave the queue. Others have a status of 'Unread' that never seems to change. During this time, no emails were being sent. Internally, externally, nothing was moving. Restarting the Postoffice service seems to have got emails flowing again but I'm confused as to why this happened in the first place. Looking at the logs nothing stands out, in fact the MTA and POC logs show that mail is flowing normally.

R4LRetro
Posts: 49
Joined: Wed Mar 25, 2020 1:32 pm

Re: Postoffice Connector Queue hangs intermittently

Post by R4LRetro »

I found the root cause.

The issue was because of a _activity.blk file in a user's mailbox. Any emails sent to this user while this .blk file existed caused them to sit in the SF queue with a "delivering" status. During this time, the user's mailbox was not accessible through Webmail or through an email client. The POC and MTA logs showed nothing pertaining to an issue.

I tried to rename the .blk file to something else and was told the IMAP service was using it. Checking the IMAP logs, I see thousands of errors for this user:

Code: Select all

WaitForAccessToFile: Blocking on activity file
WaitForAccessToFile: Blocking timed out on activity file
Error: Activity file access could not be obtained within the threshold of 100 seconds. Retrying.
It looks like the user may have set his email client to expunge or clean his Inbox upon exit. Nearly all of the errors in the IMAP log show the following:

Code: Select all

Expunge request on example.MAI. Failed to delete path\to\user's\deleted items folder\example.MAI. Error 2
I restarted the IMAP service and the .blk file disappeared. I could now log into the user's mailbox in Webmail and in an email client. Emails that were shown as "delivering" in the SF queues were now processed.

I believe part of the issue also may be that the user has a redirect rule on their mailbox. Because of this it seems the same emails failing in the Inbound SF queue were also failing in the Outbound SF queue and stopping the delivery of other emails. I'm not sure why this is happening now overall.

I've ran chkdsk and reported no errors. However I see the permissions for the Postoffices folder and beyond no longer have IME_STORE_GROUP in the Security tab. IME_ADMIN is okay, but IME_STORE_GROUP is nowhere to be found in the Security tab. Going to run ICACLS and re-set the permissions and see what happens.

R4LRetro
Posts: 49
Joined: Wed Mar 25, 2020 1:32 pm

Re: Postoffice Connector Queue hangs intermittently

Post by R4LRetro »

Wanted to update this thread.

The issue is resolved for now. Running the ICACLS commands from this MailEnable KB fixed this issue: https://www.mailenable.com/kb/content/view.asp?ID=ME020073

Since I'm on NVMe I was able to run this during off-hours and have it finish in about a half hour on 230-ish mailboxes, about 9.5million files total.

Post Reply