Charles Marcus writes:

Looking in the logs, everything is running smoothly for months, then all of a sudden, I see this:

"Jan 28 14:09:31 moria famd[6105]: segfault at 0000000000000000 rip 0000000000000000 rsp 00007fffffa0acd8 error 14
Jan 28 14:09:31 moria imapd-ssl: FAMNextEvent: Input/output error
Jan 28 14:09:31 moria imapd-ssl: FAMNextEvent: Input/output error
Jan 28 14:09:31 moria imapd-ssl: FAMNextEvent: Input/output error
Jan 28 14:09:31 moria imapd-ssl: FAMNextEvent: Input/output error
Jan 28 14:09:31 moria imapd-ssl: FAMNextEvent: Input/output error
Jan 28 14:09:31 moria imapd-ssl: FAMNextEvent: Input/output error
Jan 28 14:09:31 moria imapd-ssl: FAMNextEvent: Input/output error
Jan 28 14:09:31 moria imapd-ssl: FAMNextEvent: Input/output error"


FAM crashed.  Could be an internal bug, could be flaky hardware.


And from that point on I get a ton of these:

"Jan 28 14:09:31 moria imapd-ssl: Failed to create cache file: maildirwatch (user)
Jan 28 14:09:31 moria imapd-ssl: Error: Input/output error
Jan 28 14:09:31 moria imapd-ssl: Check for proper operation and configuration
Jan 28 14:09:31 moria imapd-ssl: of the File Access Monitor daemon (famd)."

Everything is still working fine, just these messages in the logs.

Courier has a "Plan B" when FAM breaks, which is why "everything is still working fine".

For now.


I have tried stopping/restarting courier-imap, famd, and portmap, all to no avail.

When I try to stop famd, I get the two red [!!] indicating a problem stopping it. If I try to restart it, I get the same thing, and if I try to start it, it says it is already started.

You probably have an existing hung famd process.  Find its pid and kill it.

If all fails, reboot the box.

Attachment: pgpgTUVQi00BZ.pgp
Description: PGP signature

Reply via email to