Something definately is going wrong, this cannot be a coincidence anymore.. Just had -another- disk disabled.

 

diagostics are attached..

 

First failure: http://lime-technology.com/forum/index.php?topic=55008.msg525059#msg525059

Second: http://lime-technology.com/forum/index.php?topic=55393.msg527985#msg527985

 

Bad thing is that a parity check was running and it "corrected" 64 errors... It was allmost completed, I terminated it to be sure.

 

I stopped the array but cannot unassign disk7, it already tells me it is "no device"..

 

I will now start and then stop the array again to see if I can get info on the drive and start a rebuild..

tower-diagnostics-20170129-2138.zip

I've seen a lot of similar errors lately both with SASLP and the SAS2LP, but it only appears to affect a few users, if possible I would get a different controller, one with and LSI chipset.

After starting the array again disk shows as missing and there is no unassigned disk..

 

Now doing a reboot..

I've seen a lot of similar errors lately both with SASLP and the SAS2LP, but it only appears to affect a few users, if possible I would get a different controller, one with and LSI chipset.

 

:-(  I just ordered two new saslp's for my secundairy server.. These things have been flawless for YEARS..

 

What I did before this all began was add some more 5in3 cages.. Wondering if one of them is faulty..

After the reboot disk7 is back as unassigned device, no smart values failing and/or pending sectors..

 

I started a rebuild back onto the same disk..

I recently bought a Supermicro chassis and migrated my disks and SAS2LP controller to it, everything was fine at first, then started having issues. My Supermicro SAS2LP has been used in my previous hardware with SATA breakout cables supporting 8 disks, ran fine for about a year, not one issue. Once I moved things over to the new Supermicro chassis I started having issues, so I replaced a cable, nope still had issues. Finally swapped out the SAS2LP for a Dell Perc H310 that I flashed, been running fine for two days now, not a single issue. Johnnie helped me with my issues which was great, hope you get yours sorted out.

I recently bought a Supermicro chassis and migrated my disks and SAS2LP controller to it, everything was fine at first, then started having issues. My Supermicro SAS2LP has been used in my previous hardware with SATA breakout cables supporting 8 disks, ran fine for about a year, not one issue. Once I moved things over to the new Supermicro chassis I started having issues, so I replaced a cable, nope still had issues. Finally swapped out the SAS2LP for a Dell Perc H310 that I flashed, been running fine for two days now, not a single issue. Johnnie helped me with my issues which was great, hope you get yours sorted out.

 

THanks... The SMs have been running for something like 4 to 5 years.. Its a bit of a coincidence that stuff starts going down jut after i changed chassis.. I ofcourse am now running with 4 5in3 cases... extra fans and stuff.. Could this be a power issue guys ?  Yesterday stuff went south while a parity check was running, so all disks were spinning..

My setup was super stable until I changed chassis. I had drives dropping out of the array during parity checks too.

There have been a lot of issues reported on the forums r.e. the SAS2LP's - which I've always considered excellent controllers.  I don't use any in my UnRAID systems, but have used them for friends with no issues.

 

I have to wonder if there's been some change in the driver support for these in v6 -- they seemed to have been rock-solid in v4 and v5 systems.

 

And in v6... this only started since  two months for me.

 

 

Verzonden vanaf mijn iPhone met Tapatalk

And in v6... this only started since  two months for me.

 

The problems appear to only affect some combinations of hardware and software, recent kernels seem more prone to these issues, I recently had a similar issue on my main server with a SASLP, it dropped a disk with no apparent reason, I replaced it with a LSI just in case but I'm using both the SASLP and SAS2LP on other servers with the same unRAID version without any issues so far, although those servers are only on a few hours per week.

And in v6... this only started since  two months for me.

 

The problems appear to only affect some combinations of hardware and software, recent kernels seem more prone to these issues, I recently had a similar issue on my main server with a SASLP, it dropped a disk with no apparent reason, I replaced it with a LSI just in case but I'm using both the SASLP and SAS2LP on other servers with the same unRAID version without any issues so far, although those servers are only on a few hours per week.

 

Shall I send Tom a mail ?

 

Today my 3rd SASLP arrived :-} ..  I have put it in my backup server quickly (no time for big work, I am a bit under the weather and need to get back into my bed).

 

If that now also starts dying there definately is something wrong..

 

Disk7 on the primary is allmost done doing its rebuild, when I feel better I will make sure to move the disk out of its current cage position and empty out that cage to see if it is localized to one specific cage..

One thing that one needs to be careful of the SASLP controllers is to make sure they are properly seated in the motherboard. It is very easy to have them slightly misaligned and this tends to lead to them appearing to work fine under light loads but drop out under heavy loads.

One thing that one needs to be careful of the SASLP controllers is to make sure they are properly seated in the motherboard. It is very easy to have them slightly misaligned and this tends to lead to them appearing to work fine under light loads but drop out under heavy loads.

 

I agree with this and it's something that's often overlooked. I usually have to bend the metal bracket slightly to make sure that the card still fits squarely in the slot when the screw is tightened.

 

One thing that one needs to be careful of the SASLP controllers is to make sure they are properly seated in the motherboard. It is very easy to have them slightly misaligned and this tends to lead to them appearing to work fine under light loads but drop out under heavy loads.

 

I agree with this and it's something that's often overlooked. I usually have to bend the metal bracket slightly to make sure that the card still fits squarely in the slot when the screw is tightened.

 

Would that not cause all drives on the controlles to fail though ?

One thing that one needs to be careful of the SASLP controllers is to make sure they are properly seated in the motherboard. It is very easy to have them slightly misaligned and this tends to lead to them appearing to work fine under light loads but drop out under heavy loads.

 

I agree with this and it's something that's often overlooked. I usually have to bend the metal bracket slightly to make sure that the card still fits squarely in the slot when the screw is tightened.

 

Would that not cause all drives on the controlles to fail though ?

it does tend to, but often this is not obvious until you stop the array and unRAID tells you that all the drives on the controller are now missing.

One thing that one needs to be careful of the SASLP controllers is to make sure they are properly seated in the motherboard. It is very easy to have them slightly misaligned and this tends to lead to them appearing to work fine under light loads but drop out under heavy loads.

 

I agree with this and it's something that's often overlooked. I usually have to bend the metal bracket slightly to make sure that the card still fits squarely in the slot when the screw is tightened.

 

Would that not cause all drives on the controlles to fail though ?

it does tend to, but often this is not obvious until you stop the array and unRAID tells you that all the drives on the controller are now missing.

 

In my case the array was fully functional with one simulated drive... So it does not seem likely a loose connection was the case then... I will reseat them anyhow ofcourse..

drive7 rebuild is completed succesfully.