Appreciate the help — when I finally was able to strace smtpd, I discovered that it was stalling on flock(</var/spool/postfix/pid/inet.smtp>, LOCK_EX) for some reason.
I have /var/spool/postfix as one of the persistent named volumes in the Docker configuration (so that any of the queues are not volatile in the event the container is recreated), but in this case I didn’t have a problem with deleting that volume and letting it get recreated. That apparently fixed the issue! I have a backup of that volume before I removed it so theoretically I could investigate further but I don’t know how likely I am to find anything — the inet.smtp file was zero length, wasn’t showing as locked by anything else from `lsof`, and had the same permissions as the equivalent in my test environment — the only thing different is that it has selinux attributes in the docker environment as they are passed through from the host (but they are not enforced inside the container). Well, this was an instructive exercise anyhow, and I’ve got a few new troubleshooting tools under my belt now. Thanks again! -Nicholas Reilingh _______________________________________________ Postfix-users mailing list -- postfix-users@postfix.org To unsubscribe send an email to postfix-users-le...@postfix.org
