4001 of 4039 active queue entries?

Any idea on what this means? Running Debian 4 with latest Virtualmin pro.

Nov 13 16:12:26 ns1 postfix/qmgr[2437]: warning: mail for ns1.confirming-email.com is using up 4001 of 4039 active queue entries
Nov 13 16:12:26 ns1 postfix/qmgr[2437]: warning: you may need to increase the main.cf local_destination_concurrency_limit from 2
Nov 13 16:12:26 ns1 postfix/qmgr[2437]: warning: please avoid flushing the whole queue when you have
Nov 13 16:12:26 ns1 postfix/qmgr[2437]: warning: lots of deferred mail, that is bad for performance
Nov 13 16:12:26 ns1 postfix/qmgr[2437]: warning: to turn off these warnings specify: qmgr_clog_warn_time = 0

It means you’re getting a lot of mail. :wink:

If you think you shouldn’t be, you might want to look at the mail.log to see what’s coming in. If you’re being bombarded by spam or viruses from a particular host, you might want to block it at the firewall level.

Lots of things can make mail delivery slow, though in a Virtualmin system, it is handed off to Procmail and most of the slow stuff happens there (and it isn’t subject to the same queue restrictions). You might check to be sure you have plenty of memory and enough disk bandwidth to handle all of the load being put on it (by all services). CPU is almost certainly not an issue–memory limits are generally hit long before CPU limits in modern server hardware. If you really are getting a lot of legitimate mail, and no other bounds are hurting performance, you can bump that limit up by a couple…but if you are hitting other limits, doing so will only make it worse.

I was more concerned that somebody was using my server for relaying spam. I don’t seem to be getting that much email

Actually it’s bad enough I shut the server off over night until I can figure this out. There are only 3 domains and about 9 email accounts on the whole server. It is actually using 20000 of 20000 last time I checked. The mail log doesn’t seem to show anything out of the ordinary. Any ideas on what else to check or do?

Thanks.

Mike

I have this line in the postfix config file, but on another site I saw that it shows a comma between each item. Is that correct?

Mine:
smtpd_recipient_restrictions = permit_mynetworks permit_sasl_authenticated reject_unauth_destination

ON other site:
smtpd_recipient_restrictions = permit_mynetworks,permit_sasl_authenticated,reject_unauth_destination

Joe’s right,
You want to check the mail log. What you didn’t say is if the domains were existing for some time on the server and this suddenly started or if you just put these domains on a new setup.

It’s been awhile but I’ve seen similar reaction that literally shut down lower resource servers when moving a domain from someone’s cheap windows hosting that used blackhole mail dumps for non-existing email recipients. If the spammer was using “book of names” ( essentially, adam, adam1,adam2,adam3,adam4,alec,alec1,alec2,alec3 …etc) type of brute force mailing to create mail lists those names would never return as non-user making their entire list valid. When it moved to the linux server the emails were coming in faster than you could delete them and then killing the queue when trying to bounce them.

Or you might want to check for a forwarding loop. Someone has forwarded to point B from point A and point B actually forwards back to point A because when, after returning from vacation, the user forgets to turn off the initial forward.

just some thoughts

The domains were on a server at networkredux before. I have them on the virtualmin server for approximately 2 months and this seems to have just started happening. The main reason I setup the server was to run my own email autoresponder script.

I’m the only user on the server and don’t have any email forwarding setup, except for the catch all email address.

I do see this after restarting the server:

Nov 14 09:45:04 ns1 postfix/qmgr[2436]: 2BBB15854B3: from=<confirming-email@ns1.confirming-email.com>, size=1073, nrcpt=1 (queue active)
Nov 14 09:45:04 ns1 postfix/qmgr[2436]: C61C9586622: from=<confirming-email@ns1.confirming-email.com>, size=1073, nrcpt=1 (queue active)
Nov 14 09:45:04 ns1 postfix/qmgr[2436]: C04548FABB5: from=<confirming-email@ns1.confirming-email.com>, size=1084, nrcpt=1 (queue active)
Nov 14 09:45:04 ns1 postfix/qmgr[2436]: E303E8FB98A: from=<confirming-email@ns1.confirming-email.com>, size=1073, nrcpt=1 (queue active)

confirming-email is the domain name, so I guess is the main user.

Oh I forgot to add that the server and scripts only seem to send email right after I restart it. For some reason, it stops sending after it is running for a bit. I don’t know exactly when.

Maybe some kind of bounce loop. Hard to say…Postfix usually gives clues when it sees problematic configuration, but this could be something broken in the auto-responder which might be outside of the vision of Postfix.

I’ll ping you via email. Might be best for me to have a look personally.