Lorens, Thanks for your info about the mailbox size limit. I do have log watch running and have not received any emails. Took care of that by expanding the size limit to 20MB. As for doing a grep search on 884643E30022 (The ID that said stuck in queue for 1 day), here is the output:
# egrep 884643E30022 /var/log/maillog Dec 12 10:08:50 batch-ca4-02 postfix/pickup[26935]: warning: 884643E30022: message has been queued for 1 days Dec 12 10:08:50 batch-ca4-02 postfix/pickup[26935]: 884643E30022: uid=500 from=<x...@xxxx.com> Dec 12 10:08:50 batch-ca4-02 postfix/cleanup[26937]: 884643E30022: message-id=<20111212180850.884643E30022@batch-ca4-02> Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: qmgr_active_feed: incoming/884643E30022 Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: qmgr_message_alloc: active 884643E30022 Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: 884643E30022: from=<x...@xxxx.com>, size=1503, nrcpt=1 (queue active) Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: qmgr_peer_select: 884643E30022 smtp 192.168.X.X (5 of 10) Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: qmgr_job_retire: 884643E30022 Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: send attr queue_id = 884643E30022 Dec 12 10:08:50 batch-ca4-02 postfix/smtp[26956]: input attribute value: 884643E30022 Dec 12 10:08:50 batch-ca4-02 postfix/smtp[26956]: deliver_request_get: file active/884643E30022 Dec 12 10:08:50 batch-ca4-02 postfix/smtp[26956]: 884643E30022: to=<x...@xxxx.net>, relay=192.168.X.X[192.168.X.X]:25, delay=102997, delays=102997/0/0/0.05, dsn=2.0.0, status=sent (250 2.0.0 Ok: queued as 8DBA6D0016E) Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: qmgr_active_done: 884643E30022 Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: 884643E30022: removed Dec 12 10:08:50 batch-ca4-02 postfix/qmgr[26936]: qmgr_job_free: 884643E30022 smtp Here is everything that I could find between the hours where the "read timeout on cleanup socket error" happened. (Dec 11 05:31:27 batch-ca4-02 postfix/cleanup[31691]: warning: 8A2993E3003B: read timeout on cleanup socket) Dec 11 04:04:26 batch-ca4-02 postfix/pickup[25527]: 9CC1F3E3003A: uid=0 from=<root> Dec 11 04:06:17 batch-ca4-02 postfix/smtpd[31631]: dict_eval: const pickup Dec 11 04:06:17 batch-ca4-02 postfix/smtp[31689]: dict_eval: const pickup Dec 11 04:09:19 batch-ca4-02 postfix/pickup[25527]: 7905B3E3003A: uid=500 from=<x...@xxxx.com> Dec 11 07:15:59 batch-ca4-02 postfix/smtpd[5663]: dict_eval: const pickup And here is everything before the "read timeout on cleanup socket": Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute name: diag_type Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute value: (end) Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: private/smtp socket: wanted attribute: diag_text Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute name: diag_text Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute value: (end) Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: private/smtp socket: wanted attribute: mta_type Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute name: mta_type Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute value: (end) Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: private/smtp socket: wanted attribute: mta_mname Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute name: mta_mname Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute value: (end) Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: private/smtp socket: wanted attribute: action Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute name: action Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute value: (end) Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: private/smtp socket: wanted attribute: reason Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute name: reason Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute value: (end) Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: private/smtp socket: wanted attribute: status Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute name: status Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute value: 0 Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: private/smtp socket: wanted attribute: (list terminator) Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: input attribute name: (end) Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: qmgr_queue_unthrottle: queue 192.168.X.X Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: qmgr_active_done: 177013E30037 Dec 11 04:07:59 batch-ca4-02 postfix/smtp[31689]: master_notify: status 1 Dec 11 04:07:59 batch-ca4-02 postfix/smtp[31689]: connection closed Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: 177013E30037: removed Dec 11 04:07:59 batch-ca4-02 postfix/qmgr[28082]: qmgr_job_free: 177013E30037 smtp Dec 11 04:09:19 batch-ca4-02 postfix/pickup[25527]: 7905B3E3003A: uid=500 from=<x...@xxxx.com> Dec 11 04:09:19 batch-ca4-02 postfix/cleanup[31691]: 7905B3E3003A: message-id=<20111211120919.7905B3E3003A@batch-ca4-02> Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: trigger_server_accept_fifo: trigger arrived Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: master_notify: status 0 Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: request: 87 (W) Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: qmgr_scan_start: start incoming queue scan Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: master_notify: status 1 Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: qmgr_active_feed: queue incoming Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: qmgr_active_feed: incoming/7905B3E3003A Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: qmgr_message_alloc: active 7905B3E3003A Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: 7905B3E3003A: from=<batc...@usaepay.com>, size=1084, nrcpt=1 (queue active) Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: start sorted recipient list Dec 11 04:09:19 batch-ca4-02 postfix/qmgr[28082]: qmgr_message_sort: x...@xxxx.com Dec 11 04:09:39 batch-ca4-02 postfix/smtpd[31631]: rewrite stream disconnect Dec 11 04:09:39 batch-ca4-02 postfix/smtp[31689]: idle timeout -- exiting Dec 11 04:16:47 batch-ca4-02 postfix/smtpd[31631]: idle timeout -- exiting Dec 11 05:09:19 batch-ca4-02 postfix/cleanup[31691]: warning: timeout on cleanup socket while reading input attribute name Dec 11 05:30:09 batch-ca4-02 postfix/smtpd[5663]: dict_eval: const mail Dec 11 05:31:27 batch-ca4-02 postfix/cleanup[31691]: warning: 8A2993E3003B: read timeout on cleanup socket Dec 11 05:53:51 batch-ca4-02 postfix/smtpd[5663]: dict_eval: const ipv4 I know I originally stated that the email is from the pickup but now I'm not so sure. I'm seeing a lot of postfix/smtp verbose messages. I hope this helps. Gonzo Fernandez Network Engineer On Dec 15, 2011, at 2:07 PM, Lorens Kockum wrote: > On Thu, Dec 15, 2011 at 11:49:35AM -0800, Gonzo Fernandez wrote: >> /var/log/maillog:Dec 14 04:03:07 batch-ca4-02 postfix/sendmail[12280]: >> fatal: root(0): queue file write error >> /var/log/maillog:Dec 15 12:03:07 batch-ca4-02 postfix/postdrop[21744]: >> warning: uid=0: Illegal seek > > In addition to what Wietse asked for in another mail just now, > do look at the output of the commands > > mount > > dmesg > > That should show if there is a hardware problem. > > If it is not, then is it not possible that the 8A2993E3003B you > are worrying about is a locally submitted mail from some system > daemon (like logwatch) that is too big for your configured > system limits? The times on the error messages make me suspect > some kind of cron job. > > See: > > http://www.electrictoolbox.com/logwatch-postfix-sendmail-errors/ > > > http://www.cyberciti.biz/tips/linux-unix-posttfix-mutt-illegal-seek-error.html > > 8A2993E3003B would then be a minor consequence of your main > problem, which is mail freezing up over weekends. > > To diagnose that problem, it would be interesting to grep on a > queue id that was correctly queued, like (from an earlier mail): > > Dec 12 10:08:50 batch-ca4-02 postfix/pickup[26935]: warning: 884643E30022: > message has been queued for 1 days > > Since mails such as those have been correctly queued, there > should be a wealth of information about when and how and why > it's staying in the queue so long. > > HTH >
