Areca RAID headaches

R

runnin17

[H]ard|Gawd
Joined
Feb 27, 2007
Messages
1,312
Okay, so I have an Areca 1882IX-16.

I have a RAID5 array with 4x 3TB Seagate ST3000DM001 drives. This array has been running perfectly without any issues.

I purchased 4x 3TB WD Red drives. Popped those drives in setup another RAID5 array and was off. Then about 5 minutes into this, the Areca card froze. Don't ask me how, it just froze. It would no longer respond to any ArcHTTP commands.

After a reboot I get into the ArcHTTP and I was greeted with a WTF screen. My RAID5 Seagate array had failed (disk 1 and 2 were present, disk 3 was "missing" and disk 4 was assigned to the WD Red RAID5 array).

THE PROBLEM: The card was using disk 4 of the Seagate array to rebuild the WD Red RAID5 array. SO really, WTF. This is bullshit.

I have been back and forth with Areca support. The card recently had the stock ram stick crap out on me. I replaced it and it has been working fine. However, now the card will not boot up at all if any drives are attached. Repeat, the only way the card will boot up is if no drives are attached. I know the drives are not the issue because they have been working fine. Needless to say this card is pissing me off and the fact that I may have lost some of my data due to the inability of the card to be reliable is just icing on the cake.

No real important data, I have most of it backed up. That is not the point. Areca charges a crap ton of money for these cards. They should be much more reliable.
 
Well it looks like it was my Norco 4220 backplane that was the culprit. I am still pissed that the 1882 card pretty much screwed me over on the raid array issue. Running ZAR right now and trying to recover the data. None of the drives failed and I did not write any new data onto them so the parity shouldn't be messed up.

It will just take forever to read through the drives and reconstruct the data. I guess we will see how it goes.

Sent an email to Norco about their backplane crapping out on me. They are not very easy to get a hold of and the service email is buried. I was only able to find it via a google search. You cannot get to it through their website. Kind of shitty if you ask me.
 
Unfortunately Norco backplanes die a bit too often and Areca controllers sometimes do stupid things when disks are removed/moved/attached. :(
 
runnin17 said:
THE PROBLEM: The card was using disk 4 of the Seagate array to rebuild the WD Red RAID5 array. SO really, WTF. This is bullshit.
Click to expand...

In the Areca web mgmt GUI, under System Configuration - what was "Hot Plugged Disk For Rebuilding" set to when this occurred? Disabled? Any? Blank Disk Only?
 
odditory said:
In the Areca web mgmt GUI, under System Configuration - what was "Hot Plugged Disk For Rebuilding" set to when this occurred? Disabled? Any? Blank Disk Only?
Click to expand...

I can't remember off the top of my head. What is the default setting? I have read through that setting a couple of times but I never changed it from the default.
 
Okay, then what is it set to right now? I dont recall what the default setting is, but mine is set to Disabled because I prefer to micro-manage any failures and rebuilds manually since my cards are not in a business setting where time-to-recovery is of the essence.
 
I will set it to disabled once I get the array back up and running. Right now I am running ZAR to recover the data. Like I said, nothing was written to the drives and they are all performing fine. So, I shouldn't have any issues recovering the data. It is going to take awhile though :D
 
SO, Norco got back to me. They suggested I disconnect all the molex splitters and like an idiot I did not think of this before.

Sure enough one of the molex splitter cables was fried :( That was the problem. Removed all the splitters and everything is back up working just as it should. Still running ZAR on the dropped array, but I am hopeful that I will be just fine in the end. Lesson learned.
 
You must log in or register to reply here.
Back
Top