danblee.com

Tutorials & Knowledge Base Articles for System Administrators who wear many, many hats.

Riverbed Steelhead Appliance – Replace Hard Drive

by 1 Comment

This may not apply to all Steelheads, and may not apply to later firmware versions. Here’s the machine I did this one:

  • Steelhead 2050
  • Firmware Version: 8.0.2

I logged into my Steelhead to find that the system was degraded. Clicking on the degraded link, I could see there was an obvious issue with one of the disks. I opened a ticket with Riverbed Support and they sent out a disk right away. Between support and myself, there was an instructional disconnect, which set the stage for this post.

Riverbed Tampa - Degraded

Here are some things you’ll need in order to open the ticket:

  • A system dump, FTP’d do them, which can sometimes be a pain but they require it.
  • Your Steelhead’s Serial and Firmware Version

And here’s what you’ll have to do when you have the disk.

  1. The drive is NOT hot swappable. Do not learn this the hard way. Power down your machine first, unplug it, swap out the drive, then power it back up.
  2. When the device comes back online, it will show as degraded still. You’ll need to SSH into the box to make the drive active again.
  3. The number of the disk that failed. I had four disks numbered 0,1,2,3 and disk 3 had failed.

SSH

If you’re familiar with Cisco iOS this will be cake. You’ll need to run a command to put the new drive in a state where it will begin to rebuild. Here are your commands:

  1. Open SSH, using puTTy or something similar. I didn’t have to enable SSH by default, you should be able to use the IP address or the host name if you’ve added it to DNS.
  2. Log In
  3. en (you may be asked for a password again)
  4. config interface
  5. At this point, your name should be “hostname (config) #” – The (config) means you’re going to be making changes in the right place and the # means you’re in an enabled mode to do so.
  6. raid swraid fail-disk 3 – This will show the disk has failed
  7. raid swraid add-disk 3  – This will put the disk in a position to rebuild
  8. exit
  9. exit

Here are the final commands again just because some people don’t like lists:

1
2
3
 
hostname (config) # raid swraid fail-disk 3
 

1
2
3
 
Disk 3 failed
 

1
2
3
 
hostname (config) # raid swraid add-disk 3
 

1
2
3
 
Disk 3 added to the system
 

Log back into the web interface and you should see the status as rebuilding instead of degraded and failed. It took about 12 hours for my disk to fully rebuild overnight. 250GB Disk. Eventually, if your mail is set up, you’ll get an email saying the RAID issue has cleared and you’ll be back to a healthy status.

Let me know how this goes for everyone else.

Cheers!