First time poster here!

 

Just migrated from a windows 10 setup with drive pool to Unraid. Unraid came with its learning curve but I eventually got it setup and migrated over.

 

Over the last week or so, I'm getting the "Docker Service failed to start" and I noticed its starting to hang on array startup when it comes to starting services. Anyways... I've tried rebuilding the docker.img. Not working. I'm really not sure where to go from here. Sometimes it will start if I go into the cmd line and do /etc/rc.d/rc.docker restart.

 

I've attached the Diagnostic file.

 

Any help would be appreciated from this vast community of knowledge

 

zeus-diagnostics-20180927-2325.zip

Edited September 28, 20187 yr by Brandon87

Not seeing why docker is currently failing to start, but do see a couple of issues:

 

Sep 27 21:40:31 Zeus kernel: BTRFS info (device dm-14): bdev /dev/mapper/sdc1 errs: wr 87338313, rd 58911204, flush 257624, corrupt 0, gen 0

There's a hardware problem with one of your cache devices, there are thousands of read and write errors, start by replacing both cables.

 

 

There were corruption errors detect on disk1:

 

Sep 27 21:40:11 Zeus kernel: BTRFS info (device dm-0): bdev /dev/mapper/md1 errs: wr 0, rd 0, flush 0, corrupt 261, gen 0

You should run a scrub to find the affected files, and if it happens again in the future it could indicate a hardware problem, like bad RAM.

 

 

Check your network config, this error keeps repeating:

Sep 27 21:47:11 Zeus kernel: br0: received packet on bond0 with own address as source address (addr:94:de:80:a1:c2:b1, vlan:0)

 

As for the docker, reboot and post new clean diags if it still doesn't start.

I’ll order some new cables for issue 1.

 

ive attempted a scrub on disk 1 through the GUI before and I’ve had 261 “uncorrectable errors” since I started this project. Never seen a list of files that are affected in order to pull them off and onto another disk. Where do I find that?

 

As for the network issue...I’m not sure what the issue is. I’ve bonded 5 NICs together and dont seem to have an issue with networking at all

4 minutes ago, Brandon87 said:

Never seen a list of files that are affected in order to pull them off and onto another disk. Where do I find that?

On the syslog. Tools -> System Log

 

8 minutes ago, Brandon87 said:

As for the network issue...I’m not sure what the issue is. I’ve bonded 5 NICs together and dont seem to have an issue with networking at all

Network config is not my forté, but that error isn't normal and it means there's a misconfiguration somewhere, if nothing else it will spam your log with those errors.

So after a bit of fiddling, and looking at the system more carefully, I think I know whats wrong. After you mentioned the cache throwing lots of errors I delved into that a bit more. I have all my dockers and docker.img on cache only, and since that drive thats throwing all those errors is super old (Intel 530 with 10% life left apparently) seems to be the culprit to all those docker issues. 

 

I removed the ssd, tried to preclear to see how it would handle it...and man, that thing is dead...it would process things 1/2 as fast as the second identical ssd that I barely used since I bought it. By the time I ran an erase and clear on the disk (with pre and post read) that died drive hadn't even finished the first pre-read. Needless to say, tossed it out. Using the other ssd only right now and rebuilt the docker stuff. Everything seems to be going really well again. I'll be ordering a new ssd to mirror this one now.

 

As for the network issue, I disabled that port in the settings and I'm just using my quad port intel nic card now and not the on board nic that was throwing those errors. Problem solved.

 

 

Now I'm running a scrub on Disk 1 that has all those errors to see what is affected. What exactly would I be looking for in the system log to tell me exactly what files I need to worry about? will it create an actual list for me or just show which sectors are bad?

 

Edited September 28, 20187 yr by Brandon87

It will show the affected files on the syslog similar to this:

 Oct 6 09:30:20 Test kernel: BTRFS warning (device md1): checksum error at logical 8981135360 on dev /dev/md1, sector 18000032, root 5, inode 266, offset 739328000, length 4096, links 1 (path: test/TV.SHow.S03E07.720p.HDTV.x264.mkv)

So I seem to have figured out all the issues thanks to your advise.

 

1) Cache was a 2 part problem it seems. The primary cache 1 was pretty close to dead. Swapped that to disk 2 of the cache in the same slot. Problem solved. So I ordered a new cache drive. Wouldn't recognize it when i booted. I know its not the drive now so i thought "hey why dont I switch the ports around on MB and see if that fixes it before ordering a new breakout cable" (because my cache drives and parity all run off the MB and use a breakout to my SAS cable and dont use the HBA). Well the MB must have killed a port or something because now its fine on different ports...either way, problem solved.

 

2) Replaced the SAS cable for drives 5/6/7/8 because it wasn't recognizing 2 drives (specifically 7/8). Now recognized and not dropping out randomly. Problem solved.

 

3) Scrubbed all the drives and found all corrupted files and replaced with good versions as they are all media files (thankfully). 8 in total were listed in the system log. Problem solved.

 

Now...hopefully the system holds up!

 

Thanks again johnnie.black