Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

We're running a 500GB RAID 1 on an Adaptec 2420SA - 2x500GB Hitachi Drives. It suddenly crashed and required a rebuild -- was not even booting. It's only at 66% after about a week. Very frustrated. This is the second box with the same card and drives that has had issues with the RAID/drives. Another box had this happen twice. Both machines are Linux boxes. CentOS and Fedora.

I dunno if it's the firmware -- which currently needs an update (any help doing this over PXE would be great- I have used UDA to do PXE boots in the past).

Anyway, would love to hear about experiences with this card and firmware. I thinking of going google and just using cheap boxes no raid and have server redundancy instead.

share|improve this question

That does sound buggy to me. However, with the Dell Perc controllers there is a rebuild rate setting. I would look for an Adaptec Linux utility that might allow you see this/set it (if Adapatec has the same, I imagine they would) and see if it is set really low for some reason.

Perhaps it is set to something like "only when idle".

share|improve this answer

server wouldn't boot up - so i left it. took almost 3 weeks in total to rebuild the raid 1. Finally it booted up. I just updated the firmware to version 15611 from 11564. I hope it will help.

share|improve this answer
wow -- three weeks? amazing. – Jeff Atwood May 21 '10 at 23:27
Phew, yeah! On Friday night I started rebuilding a 1TB RAID1 volume (Windows software RAID, mind you) on a live Exchange server. It was done by Saturday morning. 3 weeks.... that's just.... yeah... wow – ThatGraemeGuy May 24 '10 at 21:47
Holy I/O contention, Batman! – sysadmin1138 Jun 18 '10 at 4:37

I had a similar issue with a HP smartarray 6400 controller and a 15(?) disk raid-5 array. Rebuilding took about 2 weeks, but afterwards I noticed there is a "rebuild priority" setting which was set to low (only when array is idle). Changing that to high, and when another disk broke a year or so later, it rebuilt in, IIRC, 8 hours.

share|improve this answer

To echo Janneb and Kyle, look for a Rebuild Priority setting. I had a 5.5TB RAID5 array a while back that took 7 days to rebuild when the priority was set to low, and only 3 days when set to high. When set to high, though, our users noticed how crappy performance was. Check the BIOS and see if you can bump up the priority a notch. And then keep a real close eye on your IO Wait values to see if your drives are keeping up.

share|improve this answer

Just did it again. It's rounding the second 3-week run. Rebuild priority was set on high in Adaptec Storage Manager. I changed the RAID from RAID 1 on 2 disks to RAID 5 on 3 disks.

We're now at about 3 weeks and its finishing up. I had upgraded the controller to the newest firmware - thinking that would fix the problem. That's why I did the RAID reconfig. Adaptec shocked me again. Interesting thing is Fedora 12 won't boot while it's rebuilding (dunno why). So I leave the machine in the Adaptec RAID configuration on boot (Control-A) to watch the progress. I wonder if that slows the speed down for some reason. If so, I am not sure where else I can "park" the machine to let it run its course at a faster rate.

Either way, I figured I would share my experience. I probably will stay away from Adaptec RAID in the future because of this issue. It's a good thing this is a test machine, because nobody would accept such performance on a live server -- insane!

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.