Desktop

Ask something
Discussion Locked

Failure of a single data system

0 helpful
5 replies

by debate

Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.
Tell us how your organization is prepared for the failure of a single data system, as featured in the July 2nd Disaster Recovery e-newsletter. Does your organization have a plan in place to restore data if a server fails?

5 replies

by fmccarter

Our network people know it all. The problem is , they are the only ones who think that. We have a hudred thousand dollars in backup hardware , and our IT department won't even take the time to learn how to use it. They think because of RAID , they will never need a back up.

by jackstanton

RAID isn't necessarily a savior. We have a Sun E-450 with a 14-drive RAID-5 array. A single drive went bad and it took 2 1/2 days to rebuild the volume, thus the entire server was kaput for that long.
Although we have a L-1000 tape library for backing up all servers (NT & Unix), our IT department decided we also needed a DLT tape drive in the server (sigh).

by gull

>> RAID isn't necessarily a savior. We have a Sun E-450 with a 14-drive RAID-5 array. A single drive went bad and it took 2 1/2 days to rebuild the volume, thus the entire server was kaput for that long. <<

Sounds like you're not really running RAID-5. A single failed drive should not take down a system. You need two failed drives for that to happen.

On our Exchange Servers' Drive Arrays, we have a hot spare in each array that immediately begins rebuilding the RAID set when one drive fails... That gives us plenty of time to get a new hard drive to replace the bad one...

We'd have to see THREE drives all go bad within the same day to lose one of the Exchange servers.

I DO agree that RAID is not the only answer. Databases become corrupt even when the hardware is working perfectly... and of course, there's always user (sometimes even Admin) error that requires a restore from backup.

But as far as hardware... In the three years I've been working in this large company (was in all small companies before), I've NEVER had to restore a server completely from backup... But I have seen at least 10-15 hard drives fail...

by xina45

DATA: We currently use a data replication solution, if the primary db server fails, the secondary db server automatically (no intervention by a person) takes over, the users generally will not even notice they have been "switched". While users are on the secondary server, we try to determine the cause of the primary server failure. After the cause has been discovered & corrected, a "recovery" software package is used to "catch-up" the down server with the data that was added since it was down. If the "recovery" fails, we do use a tape-based system to back up data and can use it to do the import if necessary.

HW: Recently we suffered a disk failure. We use RAID5 and software that allows us to hot-swap the disk so, after the replacement part was sent overnight (our central hardware location is in another state)and placed in the server, it "found" the new disk and "restored" everything back to normal. In that case the database never failed, we would have never know execpt for the warning light on the server!

PERSONNEL: There are two people on call 24/7 for the system, but if help is needed for the application from the vendor they are M-F 8-5, after hours/holiday contact is via pager. Hardware is the most difficult resource for us to obtain quickly.

by mahesa.abeynayake

What was the aproximate cost of this service.
Is the the second server in the same location
What about the other servers and the data

This discussion is closed to new replies.