Hi,

 

I have a simple personal array, 2 data disks, a single parity drive, and a cache disk.

 

I had disk1 disabled , for CRC errors, it happened a month ago but since then the server was not used. As I came back from vacation yesterday i did an extended Smart test (no error apparently) decided to try and stop/dismount/start/stop/remount and rebuild the parity.

 

the parity started to rebuild but stopped immediately. I stopped the server, tried to check all cables/connections and retried the steps above. But it did the same, and even worse, disk 2 was disabled.

 

Then I did what I think was a huge mistake : i started the array with both data disks stopped.

 

And now they're showing the message : no file system and I can't remount them. 

 

I started the array in maintenance mode, but a "check filesystem status" with the parameter -nv is giving me the following message:

 

Did I lose my data ? Any method, even painful, to recover it ? I can't tell you how distressed i feel right now...

 

Diagnostics attached.

 

Thanks,

 

tower-diagnostics-20210822-2117.zip

Edited August 27, 20214 yr by CyrilB
solved now. i'll reply the final status.

Why does parity say it is external?

 

Are you trying to use USB for the array?

Hi,

I think it is because the computer I used to build this array is my previous main personal computer. It's an old Motherboard, and all SATA ports are not identical. I tried to put the Data disks and the cache disk on the same ports, and used this additional port for the parity drive. I had not noticed it was considered external to say the truth. But everything is in the tower and plugged on the motherboard.

The Data and Parity disks are new though, I bought them less than 2 years ago for this purpose, they are server-grade versions. The cache disk is an SSD drive that I repurposed.

 

I didn't think that making a server would be so sensitive to the hardware (MB, memory, PSU). Unraid disables drives for those CRC errors that I never come across in any other configuration as an issue. It had happened already 6 months ago, but after testing the drive, I had just rebuilt the parity and nothing else happened until today. I ordered new SATA cables, shielded, just in case. They arrive tomorrow.

12 minutes ago, CyrilB said:

It's an old Motherboard, and all SATA ports are not identical.

Asus does not keep the specification of old product online, but from I could gather your P5W DH Deluxe has :

 

Storage/RAID

Intel ICH7R South Bridge:
* 1 x UltraDMA 100/66/33
* 3 x Serial ATA 3.0Gb/s avec Intel® Matrix Storage Technology pour le support RAID 0, 1 et 5

Jmicron JMB363 controller
* 1 x UltraDMA 133/100/66
* 1x Serial SATA 3.0Gb/s
* 1x External Serial ATA 3.0Gb/s (ASUS SATA-on-the-Go)
* Support SATA RAID 0, 1 and JBOD (by 1x External SATA & 1x Internal SATA)

Silicon Image 4723 Hardware RAID controller (ASUS EZ-Backup)
* 2 x Serial ATA 3.0Gb/s with RAID 0, 1 support
* Support RAID 10 cross Intel® ICH7R and Silicon Image 4723 controller

 

Not sure what is the driver status for Silicon Image ?

Maybe try to stick with the 3 Intel ports and the Jmicron ? Not sure if this could help though.

13 hours ago, CyrilB said:

I had disk1 disabled , for CRC errors, it happened a month ago but since then the server was not used.

If there was nothing written to disk1 once it got disable you can do a new config and re-assign both data disks, then sync parity, both disks show a high number of CRC errors, so if you haven't' done it yet make sure to replace the SATA cables first.

Should I repair the filesystem first? I won't let me assign a disk without it right now. Or is it unnecessary if I do a new config ?

4 hours ago, ChatNoir said:

Maybe try to stick with the 3 Intel ports and the Jmicron ? Not sure if this could help though.

 

no risk to mess up more the config ?

28 minutes ago, CyrilB said:

Should I repair the filesystem first?

No point in trying to repair the emulated disks, Unraid can't emulated two disks with singe parity, after the new config they should mount, if they don't post new diags.

4 minutes ago, JorgeB said:

No point in trying to repair the emulated disks, Unraid can't emulated two disks with singe parity, after the new config they should mount, if they don't post new diags.

 

If I do that, I will still recover the data they contain ?

 

I mean, it makes sense, I'm just not sure of what I'm doing. Still learning.

Edited August 23, 20214 yr by CyrilB

49 minutes ago, CyrilB said:

no risk to mess up more the config ?

Unraid identifies the drives with their serial number, not the port on the Controller(s).

The only time it can be an issue is with RAID controllers that change the S/N. It does not seem to be the case here.

2 hours ago, CyrilB said:

If I do that, I will still recover the data they contain ?

It should, just assign the previous data disks to data slots.

Thanks, i'll wait for the new cables tomorrow and I'll give it a try. 

Hi

 

Thanks everyone for your advice it is now solved.

 

I changed the cable, did a new config with the same data disks (and indeed the data were still there), and now the parity is rebuilt and everything works like a breeze.

 

Thanks again !