yesterday morning, Yum updated my sendmail (yes, I know I probably shouldn't allow this, but I'm new to CentOS--Never had auto updates before), and now it won't work. CentOS 4.5, sendmail 8.13.1.

Basically, I can still send mail from my server, but inbound mail cycles through 26 times, and then gets rejected because of too many hops. The end of the logs look like this:

May 21 00:33:06 sendmail[29215]: l4L4X6mu029215: from=<alden@OUTSIDEDOMAIN>, size=9317, class=0, nrcpts=1, msgid=<001101c79b60$d429bfa0$5a01a8c0@Alden>, proto=ESMTP, daemon=MTA, relay=mysecondarydnsserver [IP ADDRESS]
May 21 00:33:06 sendmail[29215]: l4L4X6mu029215: to=<alden@SERVERDOMAIN>, delay=00:00:00, mailer=esmtp, pri=39317, stat=queued
May 21 00:33:06 sendmail[29213]: l4L4X1dd029201: to=<alden@SERVERDOMAIN>, delay=00:00:05, xdelay=00:00:00, mailer=esmtp, pri=128959, relay=engineno9inc.com. [209.62.15.67], dsn=2.0.0, stat=Sent (l4L4X6mu029215 Message accepted for delivery)
May 21 00:33:08 MailScanner[28794]: New Batch: Scanning 1 messages, 9904 bytes
May 21 00:33:08 MailScanner[28794]: SpamAssassin cache hit for message l4L4X6mu029215
May 21 00:33:08 MailScanner[28794]: Virus and Content Scanning: Starting
May 21 00:33:12 MailScanner[28794]: Uninfected: Delivered 1 messages
May 21 00:33:12 sendmail[29229]: l4L4XCFZ029229: from=<alden@OUTSIDEDOMAIN>, size=9675, class=0, nrcpts=1, msgid=<001101c79b60$d429bfa0$5a01a8c0@Alden>, proto=ESMTP, daemon=MTA, relay=mysecondarydnsserver [IPADDRESS]
May 21 00:33:12 sendmail[29229]: l4L4XCFZ029229: to=<alden@SERVERDOMAIN>, delay=00:00:00, mailer=esmtp, pri=39675, stat=queued
May 21 00:33:12 sendmail[29229]: l4L4XCFZ029229: SYSERR(root): Too many hops 26 (25 max): from <alden@nyc.rr.com> via mysecondarydnsserver, to <alden@SERVERDOMAIN>

Any help would be appreciated.

Thanks,
Alden

try stopping spamassasin resending a test message. I have seen this problem before with a client of mine that was running centos. i dont rememeber what the exact problem was but i remember that it was a spamassasin problem

Actually, all I had to do was stop MailScanner (which starts sendmail on its own), service start sendmail, wait a few minutes, stop sendmail, then start MailScanner. Apparently, when I start sendmail manually, it does some extra stuff (e.g., rebuilding access.db).

In any event, it works now.

Thanks!

cool im glad that worked.

CentOS does a few weird things in there distro.