R510 with a H310 RAID 5 Seven 2T Drives

Hot swap out failed drive two ST2000NM0033 (3Gb/s) with new ST2000DM006 (6Gg/s). After 22.5 happy hour rebuilding and live production the PERC H310 lost it's config. Imported config from the drives and server once again was fully functional with a flagged drive. After an hour I shut down down and rebooted (Ctrl-R) in to the PERC Utility. I manually kicked off the rebuild leaving the drive seated and 22.5 happy hours later in production the cycle repeats.

After importing the config in the PERC Utility should I:

1. Issue the rebuild un Utility but stay in the PERC Utility until it is finished eliminating OS (Lunix) interference.

2. Force the Drive Online and use a utility to correct any remaining issues with the replacement drive.

3. Back up the happy server and replace the 6 drives to match the new drive and then restore.

4. Insert the old or new drive in a different slot if the PERC Utility and adopt it in the RAID 5 family expelling the original slot.

Hi,

It sounds like there may be another drive that has an issue or the array has some corruption. Try booting to F10 and running hardware diagnostics.

Diagnostics was the first thing I ran and Drive 0 Short Self Test was unsuccessful but this was before I loaded the config back on the PERC so there was no logical drive yet. The PERC shows all drives with "S.M.A.R.T. State: No Error." With a fault tolerance of one drive which has already been marked I can't replace any other drive except the one that's failing to rebuild at the moment.

Do you have any exsperiance with the PERC CLI and would it offer any more granular analysis or control to avoid this Utility issue in handling a simple rebuild.

Thanks

I have found drive zero not part of the configuration and used import configuration again to reestablish the RAID5. I decided to rebuild drive two without leaving utilities. 1. Is there anyway to tell the progress within utilities to calculate how long this will take? 2. Is the loss of drives from the PERC configuration a sign the controller is failing? The server has dual PSU’s that are on a UPS so I wouldn’t think it was power or battery related. Thanks for anyone’s experience with this H310 PERC controller. I’m not sure I could migrate this RAID 5, if this drive doesn’t rebuild, to a different controller but if anyone knows about that path I’d appreciate the tips. Thank you

It should show a rebuild percentage but it won’t say how long it will take. It could be the controller, but usually it is just bad blocks on multiple drives.

Thanks Josh,

It does show “Progress” and some number crunching estimates 31.5 hours to completion of a 2T rebuild. I should know by 8:45 tomorrow if the 22.5 hour cycle repeats.

Does Dell recommend another controller that would recognize the RAID configuration this H310 laid down that I could then import the config to and execute another attempted to rebuild, if this attemp fails? I’m hoping for completion mainly because the OS was functional and video storage not life or death. It something doesn’t parity check then I could understand the possible issue the utility not knowing how to handle that exception. If there is a better controller in this regards it could be the answer to saving the data other than backing it up to an external drive.

Thanks

It never finished declaring Drive one failure. Oddly enough it boots after putting Drive one back online and fails to boot when marking Drive two online. I've recovering the data to an external drive and preparing the OS for install but on what and with what is the question.

What is another controller that will work in the R510 for seven new drives as a RAID 5?

The H710 or H710P are better controllers. For 7 drives you may not want to use RAID 5, since it has a higher risk of failure.