forgot one detail -- mail *is* going out (being relayed from james server to postfix server), but much slower than usual. (i.e., there is not a complete blockage of james relaying to postfix server).
> From: [EMAIL PROTECTED] > To: [email protected] > Subject: Huge message backup in spool directory > Date: Fri, 2 May 2008 03:27:25 +0000 > > > Hi all, > > I inherited a production system which uses James 2.2.0 as a backend; it has > been running incredibly well for several years (approx 5 million > emails/month), but in the past week a glut of email has begun backing up in > the spool folder and I have no idea why. I'm not sure what log fragments or > config to include, but basically this is a backend server which relays mail > to another postfix server for delivery. We have two identical servers > running James -- a primary and a failover. The primary got bogged down > (performance plummeted, mail generation extremely slow), and I noticed over > 80,000 (!) files in the var/mail/spool folder, of the following formats: > > 4D61696C313230393536323235303031312D313836353238.Repository.FileObjectStore > 4D61696C313230393536323235303031312D313836353238.Repository.FileStreamStore > > I've occasionally looked in the spool folder in the past, and there were > never more than a few dozen files/messages in there, so clearly something is > wrong. I switched over to our failover James server, which took over and > began sending out emails at the usual rate we're accustomed to, but now 3 > hours later it has begun bogging down with 27,000 files in its spool folder. > Another potentially performance-killing side effect is that the spool > directory itself is now almost 7 MB in size (ls -ldh spool), so the OS is > probably straining just to move files into & out of there. > > Since this is happening on both servers, and traffic is not greater than > usual, I suspect it may be a network issue (our postfix server is on the same > subnet), but am not sure how to get more details on this. We have not > changed james config on these servers in (literally) years, and nothing in > the past few days. > > Is there any config I can tweak in James to get more details on the reason > for the backlog? I've looked at the spoolmanager log files, which are > verbose, but there's nothing in there which looks very alarming. > > Unfortunately, upgrading to 2.3 is not an issue for us at the moment ... > > Thanks in advance, > > Andy > > _________________________________________________________________ > In a rush? Get real-time answers with Windows Live Messenger. > http://www.windowslive.com/messenger/overview.html?ocid=TXT_TAGLM_WL_Refresh_realtime_042008 _________________________________________________________________ Make i'm yours. Create a custom banner to support your cause. http://im.live.com/Messenger/IM/Contribute/Default.aspx?source=TXT_TAGHM_MSN_Make_IM_Yours
