topic LH Pro - replaced array controller, now won't boot in ProLiant Servers - Netservers

LH Pro - replaced array controller, now won't boot

Tim Slape — Wed, 28 May 2003 12:51:40 GMT

On Monday morning the server was beeping. The NetRAID Express Tools showed two of three drives failed in the boot array. I replaced the controller with an identical one from a working machine and returned to configure array (replaced no drives). I failed to erase the config from the NVRAM of the replacement controller, so ended up New Configging. All drives Onlin, all optimal, but computer will not boot - clears all POST, NetRAID info, no errors displayed, then goes to black screen with flashing cursor in upper left, hangs there. I put the original controller back in and ran New Config and now no drives show Failed, but still no boot. I'm desperate to save the contents/ configuration of the drives - no confidence in our backups. Thank you!

Re: LH Pro - replaced array controller, now won't boot

Faez Ahmed — Sat, 31 May 2003 08:05:13 GMT

Seen this problem before and I fixed it by clearing the CMOS on the server. PS, what is the raid controller....Netraid 3?

Re: LH Pro - replaced array controller, now won't boot

meloni — Wed, 04 Jun 2003 08:58:31 GMT

hello.
when you have two disk failed on a raid 5, there is no automatic way to recover. But you can do some test:
-you need first to know if both disk went off-line at the same time or if they could be put off line on differents moments (need to check logs and or see if you have alarm enable, of other ways)
-If you could see that both disk where off-line in the same time, you could check in the disk log if one of these disk have media errors, then force on-line one of the disk that have no errrors.
at that point you will be able to reboot in degraded mode, thatn rebiuld the last disk.
-If you do not know which disk went off-line as last (it is the one you should force online) then you can only try and hope in fortune, force one online and check if system boot.

Now if you have done the test you describe, clearing the config, the only way to recover data ((if you do not initialize the raid in the process)) is to recreate a new config with the exact same data as your previous config, without initializing the raid, Hoping that both disk went off-line in the same time (so that they are aligned)