http://lime-technology.com/forum/index.php?topic=32013.0

The above is my most recent post about my server and what all had been done. I bought a new seasonic power supply, locking sata cables, cleaned up the wiring, and made sure there are no loose cables.

It has been a couple of weeks with no issues. I come home today to a red ball on the parity and another disk. I have attached a syslog and screenshot of the red ball status. The server is shut down at this time because I don't know what to do.  The syslog was too big so I zipped it. I hope this works. I would appreciate it if someone would guide me as I am lost.

syslog.zip

Replacing / rerouting cables always has an element of risk.

 

I am not an expert at reading the syslogs, but I see hard resetting link messages which is usually indicative of a bad or loose cable.

 

It appears your parity disk has been dropped from the array (it is showing as removed) and the data disk with the red ball is not being found connected to the server.  Not sure if the parity disk might actually be ok but just removed from the array? You could drop down one of the drive slots to see if the parity drive is there or not.

 

I would open the server back up and recheck the connections to the drives you are having issues with. Check the power connections, the data connections, and see if they both have some common origin (for example, are they connected to a different controller than the other drives, or are they both connected to the same power splitter). Bad power splitters have been known to cause problems. Controller cards can be loose. And data cables (even locking ones) can be bad.

 

If everything looks secure and you are still not seeing those drives try replacing the data cables.

 

Good luck!

"cleaned up the wiring"

Locking cables are a great idea.

 

99% of SATA cables aren't shielded. They can have 'crosstalk', or 'radio interference' with each other. I hope 'clean up' didn't include routing them together or tying them into neat little bundles. 

I just got the new data cables and made sure they were all connected good and snug. I dont have them tied together or anything like that because another poster said they could get crosstalk. I have all the power wires snug also. I've been having this problem for the last few weeks but this time is after all new data cables and a new power supply. I can only push all the cables into the hdd's and mobo so many times before I think it is something else. I know all signs point toward a lose connection but I have verified numerous times and even replaced all cables twice now and come up with the same problem. Could it be the motherboard?

You've got clear hardware problems on those two disks...

It could be the MB...or the disk drive controller if you have an add-in SATA card of some sort.

 

Typically its cabling, but sounds like you've looked at that...(you did try rearranging the connections of drives to MB sockets to see if the failing drive's identity changed? That could indicate the problem is NOT in the drives but in the cables or MB)

 

But it could also be  Power Supply...why  did you replace your old one?

What are the characteristics of the new power supply and how did you provide power to the hard drives? (are they all daisy chained on the same supply wires?)

Try swapping the connection of working and non-working. Does the problem follow the connection or the drive? Repeat this process until the problem component is isolated.

I replaced the power supply because the one I had was a cheap package deal that came with the mobo and cpu. I The new psu is a http://www.newegg.com/Product/Product.aspx?Item=N82E16817151136 and the new data cables are http://www.amazon.com/gp/product/B00HE5VNFA/ref=oh_details_o00_s00_i00?ie=UTF8&psc=1

 

I have swapped around data cables to different sata ports on the mobo but didn't make a note of how I switched them. I do have a http://www.newegg.com/Product/Product.aspx?Item=N82E16816101358    When I started having issues I started moving the drives I've had issues with to this pci card and then I found that some of them did the same thing there. So that is why I decided new power supply.

 

Since the new psu, cables, and cleaning up everything, the server has been on for approximately 2 weeks. I did have to reboot a few times because after all of my troubles couch potato wouldn't work.

 

My next option is going to be to swap my step sons mobo and cpu. His has more sata ports anyways so it would be better for me to have it. I think the cpus are the same so it wont matter as far as performance for him. Plus no one can watch tv at the moment so I'm sure he or the wife will have issues with this idea   

 

Before I did this, I wanted to come here and get everyone's take on this idea. I don't want to do this if no one thinks it will help. I did have a set of sata cables http://www.amazon.com/gp/product/B003ZDNY02/ref=oh_details_o05_s00_i00?ie=UTF8&psc=1 that were so tight they would eventually pull themselves out of the motherboard. So I thought that possibly the ports on the mobo are kinda wore hence the connectivity loss.

 

I tried to tell my story without going in circles and confusing everyone so thanks for any help I get from here on out.

I am hoping someone will comment on my mobo plan to swap them out. Thanks.

 

 

Sent from my iPhone using Tapatalk

Well...since nobody else will chime in...

"Yes", if your SATA cables are so short they pull themselves out of the sockets, you should get longer ones!

And if its not that big of a deal to swap MB's go ahead. It seems unlikely that new issues would arise, and some of the existing issues could get cleared up.

Before you swap MB, check that you have the latest BIOS installed on the new MB.

With new cables and new power supply in the old rig, about the only other thing to test is a long MEMTEST...otherwise, I'm back to the fault being in the SATA Interface or the MB somewhere....swapping the MB seems reasonable to me.

 

Be sure that you know which drive is the parity drive and which is the cache drive...take a screen shot of the unRAID home screen. (just in case the configuration has to be rebuilt. Shouldn't need it...but better safe than sorry)

 

Thank you for your input DaleWilliams. I plan on doing this during the weekend. Great idea on the bios update.

 

 

Sent from my iPhone using Tapatalk

Of course I forgot to update the BIOS. I had a picture stored of how my drives were arranged in another thread and I verified that before shutting down the old rig. New mobo is in and upon boot I have all green balls except the parity and it shows blue. I disabled all plug ins and set my go script back to default before I booted BTW. I started the array in maintenance mode and it is currently running a parity-sync. I've done this twice before so hopefully I did it right. Wish me luck.

Cheering you on! 

Thanks Dale. I've finished the parity-sync and it seems as if everything is ok. At bed time I'm going to run another parity check. I have all of my plugins back up and I went ahead and updated from v5.0.4 to v5.0.5. So I'm on a different motherboard, cpu, and memory. All day I've been looking for something negative to happen and I did find it. The new motherboard doesnt have a serial connection for my APC battery backup. Time to research a fix.

I'm Glad its working again ...that must have been very frustrating to you.

Someone must have invented a serial-to-usb translator!

Yeah I think they have. I have a friend that is going to bring me one. It's still up and running so I hope it's good to go. I'm tired of worrying with this thing TBH. Thanks for your help with this.