2 x Parity disks and 10 x Data disks.

Few days ago 3 of my drives got red balled with tons of read errors. I've basically replaced every thing in the server other than the drives and same issue, so ordered three new drives. I'll mount the three old ones using unassigned plugin to see what data I can get off them.

Question on best way forward. 

1 - stop array

2 - Unassing bad drives.

3 - shutdown

4 - add new drives and power back up

Now, should I just use new config at this point, since I can't do a rebuild do to losing three drives? So new config would be the quickest and esiest way forward, building new parity off all my existing disks?

 

Thanks for any suggestions.

 

4 minutes ago, Whaler_99 said:

3 of my drives got red balled with tons of read errors. I've basically replaced every thing in the server other than the drives and same issue

Rebuild or New Config is the only way to re-enable disks.

 

1 minute ago, Whaler_99 said:

should I just use new config at this point, since I can't do a rebuild do to losing three drives?

It wouldn't even let you start the array unless you do New Config at that point.

 

Attach Diagnostics to your NEXT post in this thread, hopefully with the original drives still plugged.

 

Maybe nothing really wrong with the actual disks and there is some other problem.

Yeah, I figured New Config is going to be the way to go.

 

I've attached the Diag, wonder what they might say. I continued to get rad errors after every attempted fix.

I replace drive cables at least three times, swapped power supplies, swapped controllers and even swapped out the mobo/cpu for a different one in case it was some weird PCIe issue or something. I've essential swapped everything but the disks, lol.

 

Diag.zip

SMART for those 3 disks looks OK. I didn't check the others. Do any disks show SMART warnings (👎) on the Dashboard page?

 

Are you sure you don't have a power issue? Any splitters? Ideally no more than 4 drives per PSU cable.

Also, cables must have some slack, and don't bundle data cables to make things "neat".

Yeah, I ran smart, nothing, all clear. Even pulled one out and ran an extended test, fine. Like I said, I've swapped the entire PSU, I am using splitters, the 1 ->4 units, three of em, all on different PSU leads. I changed out those as well and yeah, cable management SUCKS, lol, as in nothing bundled up, etc. I heard tie up the data cables neatly can cause issues.
I'm at a loss as to what might be happening here, driving me nuts, hence the three new drives. Try that and then pull what data I can off the three old ones

4 minutes ago, Whaler_99 said:

U, I am using splitters, the 1 ->4 units,

These are normally OK if Molex->SATA but tend to be problematical if SATA->SATA.

Interesting. They are indeed SATA to SATA ones. 
I'll check how many Molex to Sata ones I have and maybe swap out to those.

16 minutes ago, Whaler_99 said:

Interesting. They are indeed SATA to SATA ones. 
I'll check how many Molex to Sata ones I have and maybe swap out to those.

Don't have any, ordered two sets of Molex to Sata, they will be here tomorrow, so I will swap those along with the drives. Might as well use the new ones. Thanks for the suggestion on that, would never have though of that piece in particular.

 

So, good news. I got two Moelx to Sata power splitters, hooked them up and the last two drives connected direct to PSU without any splitters.

Rather than adding in the three new drives, I did a new config with existing and started the rebuild and... no errors and everything been going fine. Going fine for an hour now, usually I got the errors and red ball withing 1 to 10 mins. 

Thanks @trurl and @itimpi for the suggestion. This looks to have resolved it.