Sebastien Roy wrote: > On Fri, 2008-10-31 at 09:05 -0700, Niall Power wrote: > >> Hi Seb, >> >> This is sounds quite odd and I haven't seen anything like this before. >> If your disk was extremely full (I mean more than 95% full then they >> would get deleted almost as quickly as they were created in attempt >> to keep the system usable. Is your zpool almost full? If time slider deletes >> extra snapshots it will pop up a warning on your desktop and log to >> syslog. Check for time-slider-cleanup output in /var/adm/messages >> > > That's not the issue, my zpool is happy: > > seb:~$ zpool list > NAME SIZE USED AVAIL CAP HEALTH ALTROOT > rpool 117G 26.0G 91.0G 22% ONLINE - > > >> Can you also check the following smf service logs: >> ls /var/svc/log/system-filesystem-zfs-auto-snapshot\:frequent.log >> ls /var/svc/log/system-filesystem-zfs-auto-snapshot\:hourly.log. >> ..to see if there are any problems reported. >> > > svcadm: svc:/system/filesystem/zfs/auto-snapshot:hourly: Permission denied. > Warning - Invalid account: 'zfssnap' not allowed to execute cronjobs > Warning - Invalid account: 'zfssnap' not allowed to execute cronjobs > cron may not be running - call your system administrator > > The cron service is online and the cron daemon is running. > This issue was encountered by another user today. It seems that an IPS action adds a role without a supplied password as a locked account.
Tim Foster has been looking into this on indiana-discuss: http://mail.opensolaris.org/pipermail/indiana-discuss/2008-October/013174.html Niall. > -Seb > > >
