I'm in need of some advice regarding a hardware setup I'm dealing with. I've got a RAID5 array made up of 5 1.2TB SAS drives running on an older Poweredge R540 with a PERC H740P RAID controller. Recently, one of the drives has started showing SMART errors and has been marked as a predictive failure, though it's still operational for the moment. Thankfully, I have an identical 1.2TB SAS available that's set as a global hot spare, but it's not specifically allocated to the RAID5 array.
My concern is about the best approach to replace the failing drive without risking the stability of the array during the rebuild process. I understand that just pulling the failing drive could lead to complications, and it seems like the correct procedure would involve using something like PERCCLI to mark the drive for replacement in order to safely switch over to the hot spare. However, the interface via iDRAC doesn't show options for this. Is using PERCCLI the right move, or have others successfully just swapped the drives directly? I'd love to hear how people have handled this situation before!
2 Answers
From my experience, just pull the failing drive and insert the new one. The server is designed for hot-swapping, so it should take care of the rebuild process itself. Just keep an eye on performance while it re-silvering—things might get sluggish for a bit. Always double-check with Dell support for peace of mind, and make sure your backups are solid before proceeding!
I've done a similar swap before but wouldn't recommend it without confirming the best approach first. Yanking the drive without preparation can sometimes lead to a bad situation; there’s always a chance the array could fail during the rebuild. If you're not on a support plan, you should at least check Dell's documentation. Best to be careful!
Haha, I had a colleague panic in a similar situation—things could get messy if you’re not careful about which drive you’re pulling!
Absolutely, I hear you! It’s all about being cautious, especially given the age of the server and the lack of support. Luckily, the drive is still functioning for now, so I have a bit of time to plan this out.