I have a local network with a half dozen machines on it. One serves as a firewall but also runs a mail gateway an an http proxy for the other machines on the network. Incoming and outgoing email have been working with no problems for some time (and yes, I have fixed the line in /etc/sendmail.cf ...). Recently we had a long-term power outage and the ups did not politely tell all the machines it was about to die when it went down, so everything went down hard. Rebooting, I was asked whether to run fsck on all the linux boxen (2 rh9 workstations and 1 8.0 machine which is the firewall), and I did so -- no problems were reported. I should mention that my main workstation has 2 other external connections besides the one through the firewall box to the internet: my office adds 2 other connections (both through industrial-strength firewalls). Ever since the power failure, I have been unable to send email from my main workstation either to or through the firewall machine, OR to anywhere else through its other connections. This box is running 9 and was sending and receiving mail just fine through all 3 (depending on destination) until the outage. I have checked the /etc/mail files (sendmail.cf, etc.) against the backed up copies I had made on another partition the day before the power failure and they are identical, as are the configurations for portsentry, networking, etc. ... virtually every other config file remains unaltered, and a check (md5sum) of the files under /bin, /sbin, /etc, /usr/bin, and /usr/sbin are also identical (they were all backed up at the same time); permissions and ownerships remain unchanged as well. FWIW, I have also swapped out all of the boards in the machine with others here, and have gone so far as to move the drives from this machine to another, similar machine as a test -- all with the results unchanged. When email arrives at this machine, here is a typical indicator in /var/log/maillog: > Jun 30 23:57:25 34 sendmail[7103]: h613vPUQ007103: tcpwrappers (AAAAA, ###.###.###.###) rejection where AAAAA was the machine name and ###... was the ip address (nothing wierd shows up in the log for portsentry, BTW). *Occasionally* a piece of email will be ACCEPTED for delivery on this machine (from an external machine) but it is never delivered. Most of the time (but not always) I am still able to send mail from this machine; but the times when it fails do not follow any distinct pattern. Internal email is unaffected. On this machine I am running : > kernel-2.4.20-18.9 > glibc-2.3.2-27.9 > sendmail-8.12.8-4 Any suggestions (what to look for, what obvious stuff I have overlooked in RTFM'ing, etc.) would be greatly appreciated. Bill Austin
