FQ Alert: PHOENIX

[Deb]

Hrrmmmm... say something nice about a server and.....

PHOENIX is not wanting to come back up via remote reboot.[nbsp][nbsp]Andrew is on his way now to bring the server back online.

ETA less than 60 minutes...

Deb

[Deb]

Everything's back up... Andrew mentioned a failed drive... I'm sure he'll make a post to explain once he's out of the chill factory and back in the office....

Deb

[Terra]

The PHOENIX server had a SCSI hard drive failure that would not allow the server to be restarted...

The RAID system + the new ReiserFS system did their job beyond all expectations...[nbsp][nbsp]I dug through the log files from some of the busiest sites on the server and I did not find any corruption whatsoever...

It appears that the new setup is going to pay off with some peace of mind knowing that the servers are protected from 3 different angles...

1) RAID 1 setup
*prevents lengthy downtimes from restoration from backups
2) ReiserFS Journaling
*avoids long e2fsck (ScanDisk) on bootup, was: 18 min // now: <2 min
*makes sure that dirty buffers are preserved and written to drive from the journal upon crash startup and recovery
*cutting edge technology that appears to be quite stable even in the event of disk failure
3) Backside backups
*in case both of the above fail and used for Armageddon restorations only

I had hoped to never test the system on a live production server, but I can honestly say that the core architecture has exceeded even my expectations... :)

--
Terra
--Another day in the life of a sleepless sysAdmin--
FutureQuest