Server Outage

[Grimjaw]

depending on the size of the array, and the raid configuration (1,4,5,etc), usually 1 dead disk wont take down the server. it would take multiple disks to fail before it would take down the server, which gives you a chance to replace failed disks before it has an impact.

[Wisteso]

Originally Posted by Grimjaw [You must be logged in to view images. Log in or Register.]

depending on the size of the array, and the raid configuration (1,4,5,etc), usually 1 dead disk wont take down the server. it would take multiple disks to fail before it would take down the server, which gives you a chance to replace failed disks before it has an impact.

Yep. With raid4/5, one bad disk causes degraded performance but not a total outage.

My guess is the card itself died since the data is still intact according to Rogean. New cards range from 10 to 3000 dollars. #donate

http://www.newegg.com/Product/Produc...82E16816124009 low end
http://www.newegg.com/Product/Produc...9SIA67S2FB5169 higher end
http://www.newegg.com/Product/Produc...9SIA67S2985740 enterprise end

If he can find an exact replacement, should be back up tonight. If he needs a different model, might take half a day.

[giesler79]

Raid array isnt the issue from Rogean (sp) post. Its the DB server. Swapping in a new disk into a raid 5 setup is trivial. May wanna just call it a night and log in tomorrow =)

[Wisteso]

Originally Posted by giesler79 [You must be logged in to view images. Log in or Register.]

Raid array isnt the issue from Rogean (sp) post. Its the DB server. Swapping in a new disk into a raid 5 setup is trivial. May wanna just call it a night and log in tomorrow =)

Originally Posted by Rogean [You must be logged in to view images. Log in or Register.]

It is an array failure, but nothing is lost. Not even a rollback (We have active replication). But we need to get the database server back up and running. Could be a couple hours folks.

If he's relying on replication, I guess that means the disks suffered unrecoverable corruption? Not using raid 4/5 I guess, if they have replication? The odds of two disks failing together are pretty small.

[Shift]

I just donated 40 bones to you gentlemen of Project 1999 we believe in you.

God Bless

[President]

Originally Posted by Wisteso [You must be logged in to view images. Log in or Register.]

If he's relying on replication, I guess that means the disks suffered unrecoverable corruption? Not using raid 4/5 I guess, if they have replication? The odds of two disks failing together are pretty small.

Replication means one database server is reading the logs of another database server and replicating itself to be the same, usually a handful of seconds behind. That should mean they have a database server still good to go, but it can be a lot of work to bring an application up on the replicated database, plus the risks associated.

[Saker]

Originally Posted by giesler79 [You must be logged in to view images. Log in or Register.]

Raid array isnt the issue from Rogean (sp) post. Its the DB server. Swapping in a new disk into a raid 5 setup is trivial. May wanna just call it a night and log in tomorrow =)

You're clearly insane!

[You must be logged in to view images. Log in or Register.]

[flatpick13]

I appreciate the updates and hard work. Cyber high-five!

[mattrc]

Or it could be like the server where I work where nobody noticed the nagging alert that a disk had failed in a raid 5 array until a second disk in that array also failed. (about 1 year after the first failure)

However, since there's no data loss I'll go with the 'failed card' guess.

[Grimjaw]

Originally Posted by mattrc [You must be logged in to view images. Log in or Register.]

Or it could be like the server where I work where nobody noticed the nagging alert that a disk had failed in a raid 5 array until a second disk in that array also failed. (about 1 year after the first failure)

lol so true! damn infrastructure teams!