Xqm Runs In Infinite Loop Of Query_status () ?

[ciuly]

I have a long time running xmail setup and I figured I'd take a peek at the spool since the folder is taking about 4 GB disk space. So XQM seems to be the tool for it.

I installed the agent on the linux machine with xmail, then the windows client and configured everything up and connected. Then clicked on the "query" button and since the, it just "hangs" (it's not frozen).

I started the agent in debug mode from beginning and all I see after query click is
<date> <time> <my win IP> : query_status ()
every couple of seconds.

I let it run for over an hour, still nothing. I will leave it running till the evening, just in case.

So .. what to look for?

note: On the linux machine, the xqmagent is eating up the CPU (running at 4 threads)
Also,
find . -type f -print | wc -l
returns 918405 (yes, 1 mil). I have a feeling this would be the culprit, although since my normal fcount and searching takes seconds to process, I don't see why xqmagent would take hours to process the files.
find . -type f -print | grep "/froz/" | wc -l
returns 458932
find . -type f -print | grep "/slog/" | wc -l
returns 458936

in case the issue is indeed a huge number of frozen messages, can't the xqm be made in such a way that it works on batches or something so it's more responsive?

le: 8 hours now, still running....

[ciuly]

so this morning I see that the process finished after almost 10 hours.
But, the grid is empty. (see screenshot).

so what now?

Attached Image (Click thumbnail to expand)

[hschneider]

4GB is too much to handle. You should

- Stop XMail
- Delete all inside spool, except 'local' and 'temp'.
- Start XMail

Then check the new created spool after a while to find the reason for this.

[ciuly]

Thanks for the info.

I cleaned out and it's been running for almost a day now, with pretty much nothing left in the spool. QSize is reported to be 0.42 MB
du -hs
says it's 20 MB but AFAICS, these are only the folders with their 4kb report

I'll keep watching for a few days and if anything pops up, I'll report back.

So I inspected the backed up old spool. using the -mtime for the find, I managed to narrow down the frozen messages to have stopped about 184 days ago.
well, I think I'll investigate this deeper when I will have some time. I have the logs (since beginning) and the spooled files, so I'll check a few to see the cause, but since they are over a half year old, I don't think they make sense now to be sent out. I'll just have to somehow filter out everything newsletter, automatic emails and that sort of junk to see if any important or relevant email was missed.

Long story short, thanks for the help

[ciuly]

just as a small update.
every now and then I get a few emails which are reported with some DNS failure and these just disappear after a while. So no frozen messages. There could have been some error in the setup that it gotten so many frozen messages a while back.

So all is good.