Blade Watch

Got an email the other day about replacing the hard drive in a HP DL380. “What’s best practice?” I got asked. The process should be similar, the overall rule though is if in doubt contact your hardware support vendor/manufacturer before doing anything.

Basically the rule is that anything with purple tabs is hot swap, so you can swap the SCSI hard drives whilst the server is in use providing it has RAID setup. Let’s do the best practice steps:

• Check the server event logs for errors – it should state which drive has failed and which slot
• Check the HP insight agents – http://servername:2301 – log in and confirm this by selecting the array controller, the drive should go orange and say failed
• Check the front of the server, the drive should go orange
• Check you have a valid backup – and consider notifying users if the server is a high impact one – if it’s that high profile or you’re unsure do it out of hours
• Push the purple tab releasing the drive handle and then pull the handle to remove the hard drive
• Now, get the replacement drive (confirming it’s a compatible or identical drive)
• Plug the drive in, holding the handle and use the handle to guide in and secure the drive, it should close and click into the purple tab

The event log should show something like:

1. Drive 0 removed
2. Drive 0 detected
3. Drive 0 status has changed to active
4. Drive 0 status has changed to rebuilding (state 7 – I think)
5. Drive 0 is now rebuilding
6. Now the front of the drive should start rebuilding, the disk icon should flash. You can monitor the rebuild status using the HP Array controller or the HP insight agents.

Consider the following:

Replacing the drive should be a non-impacting event and 99 times out of 100 it is, but if the server has a fault with the array controller, it is generally unhealthy or unhappy, swapping that disk could cause further issues so do ensure that you have a valid backup. Ensuring the server and array controller firmware and drivers are up to date should avoid this.

The process above should work for any of the Compaq/HP Proliant servers with hot swap SCSI disks.

If you wish to change the drive offline – that is switching the server off. You need to simply shut the server down, wait a few seconds for the drives to stop spinning, remove the faulty drive, and power up the server, once the server has started read the F1/F10 prompts carefully selecting the wrong option can cause data loss. I normally get the server to boot into windows once the log in screen has appeared and the server has settled down and then replace the drive, but it’s up to you.

Always speak with HP before doing anything if in doubt and keep in mind some array controllers have minimum hard drive firmware revision levels for functionality or to ensure best performance.

Related posts:

1. How do I tell if my server hard drive is rebuilding I got an email from Shirley: “How can I tell...
2. Is my disk/array rebuilding on DL380? I got an email from Mike asking about his Compaq...
3. Best practice/methodology from the vendors I got an email asking if I had any plans/emails/notes...
4. Using the Compaq/HP Array controller to configure array A friend phoned up with a question last Saturday afternoon:...
5. Cisco server and blade firmware and drivers Cisco Cisco Rack Server firmware: (UCS C250 to follow, once...

Related posts brought to you by Yet Another Related Posts Plugin.