Hasn't been a good week for me.

So this all started on Sept 23rd when I was away and disk 3 failed (the server was frozen). Upon reboot I had disk 3 (WD 4TB RED) showing "Device is disabled, content emulated"

 

When I returned to work on the server. I removed the drive and installed another WD 4TB Red. The drive was detected, and added, and rebuild began (approx 20-22 hours). During that parity, the device went yellow, and contents was emulated. (sigh) 
Parity finished rebuild

So reading some more online, I read about cabling and my SAS card. 

I changed out the breakout cable
+
moved the SAS card to another PCIe16X

Again I started the server and ran a parity check. Everything checked, and was 100% successful. I rebooted the server and it came back online with no concerns. I then started to  write data to my array etc.
During the copy I found 'taracopy' say network device (unraid) unassailable. I went back to the WebGUI and refreshed and then saw another error on Disk 3

"Device is disabled, content emulated"

The device had a RED X and I checked out the Dashboard, and it was also FAILING S.M.A.R.T.
I checked the .log file and there is a ton of write/sector errors.

 

Keep in mind both of these WD 4TB REDs were precleared prior to use (from brand new) 
Now I have another WD 4TB RED sealed, and may pre-clear it before adding (takes 48 hours) 

 

But what I am trying to determine if it is actually the drives itself or the SAS card [ CONTROLLER: SUPERMICRO AOC-SASLP-MV8 ]  or something else. Very sad and time consuming, need to get back and operational. I was looking if there was any assistance. 

 

 

syslog posted below
 

syslog.txt

Edited September 27, 20178 yr by bombz

SASLP is not recommended for unRAID v6, LSI HBAs are, but it won't cause SMART issues on the disk, if SMART is failing the disk is failing.

 

Did the first replaced disk also had SMART issues or it was just disable?

The first disk replaced was JUST disabled. There was NO S.M.A.R.T issues with the first disk.
I am currently pre-clearing a BRAND NEW WD RED, just broke it open and pre-clearing on another workstation/PC

What would you recommend going forward. I have no problem purchasing a better, or recommended SAS card. I bought this back in the version 4 or 5 era. I would love a link so I can look it up. 

 

I still have a spare on-board SATA port I can use. Should I try the disk (replacement disk) on that SATA port? 
I also have a spare SATA cable on my breakout cables (on the other SAS port) Should I attempt that.
Should I be buying a new card ?

 

LSI 9300-8i PCI-Express 3.0 SATA / SAS 8-Port SAS3 12Gb/s HBA - Single--Avago Technologies

https://www.newegg.ca/Product/Product.aspx?Item=N82E16816118217

 

LSI Internal SATA/SAS 9211-8i 6Gb/s PCI-Express 2.0 RAID Controller Card, Single--Avago Technologies

https://www.newegg.com/Product/Product.aspx?Item=N82E16816118112

 

SAS9211-8I 8PORT Int 6GB Sata+sas Pcie 2.0

https://www.amazon.ca/SAS9211-8I-8PORT-Int-Sata-Pcie/dp/B002RL8I7M

 

Can I use my same breakout cables with this.
Are we sure this is the SAS card posing the concerns?
Should I be buying a new SAS card?

Edited September 27, 20178 yr by bombz

The 2nd failure was the disk, not the controller, do you have diagnostics and/or SMART from the 1st? Without them can't say what happened.

Still the SASLP is not recommended because of its low performance and tendency to drop disks without reason, both LSI you linked will work though most look for server pulls on ebay where they are significantly cheaper.

OK
I might have diagnostics sec

what about the 3rd link I added?

here are my old syspulls and diag

 

 

unraid-diagnostics-20170925-0045.zip

unraid-diagnostics-20170925-1055-after-complete.zip

unraid-syslog-20170925-0044.zip

unraid-syslog-20170925-1054-after-complete.zip

what about the 3rd link I added?

It's the same as the 2nd.

Also, you can use your current cables.

7 minutes ago, johnnie.black said:

 

 

It's the same as the 2nd.

Also, you can use your current cables.
 

OK, ordering...

SAS9211-8I 8PORT Int 6GB Sata+sas Pcie 2.0

https://www.amazon.ca/SAS9211-8I-8PORT-Int-Sata-Pcie/dp/B002RL8I7M

 

There is nothing worse then having the incorrect hardware for an important job
(sad) Standard Shipping — get it Oct. 26-Nov. 8

 

 

Edited September 27, 20178 yr by bombz

39 minutes ago, johnnie.black said:

I may take a chance, and toss the other DISK 3 back in (the first one) Which does have data on it, and hook it to onboard SATA and see if it works
What's your thoughts on that ? 
 

I am going to head to bed, get a fresh start on this tomorrow. Thank you for the assistance so far, and let me know your thoughts

for now I am keeping the server online, spun down until feedback is provided. Kind of learning on the fly with this new concern

Thank you for all the help !!

 

Edited September 27, 20178 yr by bombz

Check the SMART report first, and you'll still need to rebuild it.

OK if SMART passes etc, and I rebuild it, and we still have a concern, i wonder what the next step is then 

 

Posted is the current disk status
 

WDC_WD40EFRX-68WT0N0_WD-WCC4E0052905-20170927-1132.txt

Edited September 27, 20178 yr by bombz

updated sig as UnRaid version = unRAID Server Pro version: unRAIDServer-6.3.5-x86_64 Pro

Here is the "original" disk3 that I pulled out initially 

WDC_WD40EFRX-68WT0N0_WD-WCC4EHXH65U4-20170927-1151.txt

***********************
Replacement disk installed.
Start will start Parity-Sync and/or Data-Rebuild. 
Yes I want to do this

****************************

I am doing this now, it is going to take a good 20 hours 

 

I started this process, and DISK3 has a orange triangle next to it. Saying...  device contents emulated
Also, this disk is NO LONGER plugged into the SuperMicro SAS card, rather the Marvell on-board SATA port / chipset

 

I have also posted NEW syslog.txt for any other info that may be needed

syslog.txt

Edited September 27, 20178 yr by bombz

I started this process, and DISK3 has a orange triangle next to it. Saying...  device contents emulated

 

This is normal

 

Also, this disk is NO LONGER plugged into the SuperMicro SAS card, rather the Marvell on-board SATA port / chipset

 

Supermicro SASLP is also Marvell.

 

 

OK good to know. it's plugged into on-board (fingers crossed!) 

did you see anything else in my logs that are cause for concern by chance? 

5 minutes ago, bombz said:

did you see anything else in my logs that are cause for concern by chance? 

 

Nope, SMART for old disk looks good, and nothing of concern in the logs so far, well, maybe all your disks still being reiser, you should consider converting to xfs, reiser is dead.

OK that's good. Thanks for looking at everything!

 

Currently 
Parity-Sync / Data-Rebuild 18.5 % 

So I hope everything goes well!

I would love to move to the new file system! I wonder how much of a pain that is to do being there is data across all drives? I am sure there is a guide or tutorial ? Am I at risk of data loss converting ? 

https://forums.lime-technology.com/topic/54769-format-xfs-on-replacement-drive-convert-from-rfs-to-xfs/

 

Thanks,
I was reading that too :-)

That's a big task! Is there a disadvantage to staying with RFS ? 

Is there a disadvantage to staying with RFS ? 

 

Terrible performance in some cases, e.g., it can take minutes before a copy begins to an almost full disk.

Some users have the system go unresponsive after a few days/weeks.

It's not being actively developed/maintained.

 

 

 

That's exactly what I am experiencing 
When I do a file copy from Windows system to my array, it takes a while for it to fire up, and even then so, takes time during the copy to complete.

I want to switch over, but that's a very very long tedious task it seems
Reading about it, it seems clear and confusing at the same time 

That may be a task for another day, month, year. I just stuck with it when I started using Unraid back in version 3 or 4 (when I purchased it) 

So far so good

 

Some .log line errors I noticed are concerning me

 

Sep 27 11:49:38 UnRAID kernel: sas: --- Exit sas_scsi_recover_host: busy: 0 failed: 0 tries: 1

Sep 27 11:49:38 UnRAID kernel: sas: Enter sas_scsi_recover_host busy: 0 failed: 0

Sep 27 11:50:05 UnRAID avahi-daemon[2584]: WARNING: No NSS support for mDNS detected, consider installing nss-mdns!

Sep 27 11:50:57 UnRAID root: error: webGui/include/Boot.php: wrong csrf_token

 

SCSI is prob relating to some type of controller or chipset ? 

New screen

 

64% complete rebuild/sync

I will test for a week and close off this thread.
So far it seems stable.

It will be weeks until I get my LSI in the mail, and I will pull out that supermicro card. It's a shame I cannot use the supermicro anymore. Perhaps I will keep it around on stand-by in case I ever run into SAS card failure. Seems SAS cards take forever to get / ship in .ca