also sprach GSR <gsr.b...@infernal-iceberg.com> [2010.05.09.0601 +0200]: > I got the same email. But I think I understand why it does that and > have a fix so it stops causing false alarms. It happens because the > cron command tries to run once per month, but only the first sunday > should checks be performed. So the second test in pipeline fails > around 3 weeks per month, ending the pipe with error, when in reality > that error is not a problem. Plausible fix: > > [ -x /usr/share/mdadm/checkarray ] && [ $(date +\%d) -gt 7 ] || > /usr/share/mdadm/checkarray --cron --all --quiet > > This checks for command availability (as before) and then checks if > day is not in 1-7 (le becomes gt), so for 8 and after it will stop > there with exit code 0 (cron is happy). If that fails (&& becomes ||) > it executes the check and cron gets whatever error code the vital > command returns.
This is a good fix, but I just wonder why this suddenly happens. The cronjob has been in place for years, and I have never seen such an e-mail. -- .''`. martin f. krafft <madd...@d.o> Related projects: : :' : proud Debian developer http://debiansystem.info `. `'` http://people.debian.org/~madduck http://vcs-pkg.org `- Debian - when you have better things to do than fixing systems
digital_signature_gpg.asc
Description: Digital signature (see http://martin-krafft.net/gpg/)
