Where'd my volume go!

[Jeff Graves]

I built a replacement machine for my home/office. Half of the storage is for personal and other half is for business.

Sys info
Intel S3200SH (MB)/ Xeon 3.0 / 6 GB RAM

Storage
1ct 3 TB HST NAS HD operating stand alone (purchased brand new in march - has no issues)
2ct 1 TB build as RAID1 and is currently not reporting to the sys correctly.

The NAS crashed hard a few days ago, I heard it reboot and noted by the sounds it did not start however POST was ok. It would not start do to boot device issue, I re-installed yesterday evening and it only partially came back to life. The 3 TB container came online however the RAID array did not. Errors returned was in reference to unable to determine size of available space. In GUI the volume is not listed however in shell i can witness it is there. All shell commands appear to be able to communicate with the container. CFS share was dumped prior to the detach and is still not configured. zpool status is "OK" both drives are listed and report as "OK"

Last thing I've done is detached and reattached, this would only work tho from shell with -f command. Currently running a scrub from shell and it is 50%. Of course the RAID vol is for my business, backups of images and client info. I'm sure there are similar threads out for this condition however I cannot look through all of them.

any suggestions?

JG

 

[cyberjock]

Can you provide a debug file? System -> Advanced?

 

[Jeff Graves]

as requested.

 

[cyberjock]

RAIDBANK00 has not been attempted to be mounted since April 28th at 10:33AM, with one exception. You manually mounted the volume on April 30th at 10:21AM, then started a scrub about 18 minutes later. So RAIDBANK00 will not be accessible to the WebGUI, nor can you share it or anything (which you know about, obviously).

So, a few questions:

1. When you did a reinstall, did you restore your config file from backup to FreeNAS? If not...
2. Did you follow the manual for importing a zpool?

It sounds like you've already done one or the other, just confirming.

Also you have 6GB of RAM. We know, from experience that doesn't always work "right". And to be frank, you aren't going to find much sympathy for more help if you can't meet the minimum requirements. *I* put the 8GB in the manual because we know FreeNAS is erratic. It can work fine today and tomorrow not work. So claiming it has worked before is no defense to ignore the minimum requirement.

You clearly aren't using server grade, but that is not likely the problem here. But the hardware you are using is woefully inadequate for FreeNAS. Our hardware recommendations already put your hardware beyond what we'd recommend or test, so you are pretty much on your own if its some kind of hardware inadequacy or incompatibility.

 

[Jeff Graves]

(hardware comments) noted and noted. I have been getting better parts I wont even mention what i was using before out of fear of being shot.

to be sure your concern here is the ram not being 8GB (I read that before I set up and chose to continue) and the fact that the 1 tb drives are desktop class and not enterprise or NAS class? or is this based on the entire system as a whole?

Q1 no, it was a roll up during install and not a config import after install was completed.
Q2 yes, however the only thing i did not do is reset to factory defaults and try again. - I can do this tho

I read about a procedure to mount the drive(s) and run file copes through the shell to another container. any experience in this space, is this even possible? Or any other recommendations to at least attempt to get the data off without an over priced data recovery service? (or if i need to get more ram and then come back to the table (again if that was the only concern))

or what if i was able to build a parallel system and attempt to import and copy?

I'm sorry about the questions id rather get through it all up front instead of some of the other cases where people talk for miles about nothing.

 

[mjws00]

Less than 8GB is the primary concern. You get that erratic behaviour cyber has mentioned a thousand times. As for drives, desktop drives are often missing TLER, WD Greens need a park tweak to survive. But we run RAID(Z) so that the inevitable disk failure is planned for. They all fail enterprise, nas, desktop. Cyber can expand all he wants.

The good news is things don't appear to have gone terribly sideways yet. Plus this is a mirror, not z1 or a stripe.

You may get away with 'zpool export raidbank00' and then an auto-import in the GUI. Wait until the scrub is done... and give it a shot. Good chance of success, and still lots of options left.

You can certainly move the mirrored pool to a parallel system and import. But it really doesn't look that bad.

Of course, auto-import could also puke and crash, then we might be talking about RAM... but that pool is currently live, just not in the GUI.

 

[Jeff Graves]

It's back to life

Gold stars for everyone!

Thank you for the suggestion and info

JG

 

[cyberjock]

And he'll be back next week with some weird bizarre behavior that nobody else has ever seen before.. nor will ever see again. ;)

 

[Jeff Graves]

I put the extra ram in.

thanks again,

JG