(SOLVED) Issues after cache SSD ran out of space

[Mesias]

Hi there,

I'm not sure how to best describe the issue I'm experiencing so I'll make a list of what happened:

1. I had Deluge downloading very large files, all working as expected
2. The temporary download folder in the cache drive reached its limit after a while. Basically the single SSD I had at that time was filled to the brim.

3. I ignored the system for a couple of days which resulted in all docker apps crashing. They were all on red and the Fix Common Problems plugin was showing some error with the docker image.

4. I added a new SSD to the system, which was recognized immediately, and a cache pool was created. This didn't resolve the problem, all errors continued. No docker apps available, no updates possible.

5. Thinking the image docker was corrupted I deleted docker.img and added one app, Deluge. Deluge worked just fine and I could install maybe one more app but after a while I couldn't install anything else. The docker app install ended every time with "Error: open /var/lib/docker/tmp/GetImageBlobXXXXXXXXX: read-only file system". One of the files still downloading in Deluge showed an "Input/output error" error.

6. I've been reading the forums for a couple of days but all I've been able to find is related to a full docker.img file which is not the case. Deluge is currently only using 7% of the file. What else could be causing this issue?

Attached you can find the diagnostics folder. mesias-diagnostics-20201008-2137.zip

Thanks in advance.

[JorgeB]

Adding a new cache device by default results in a raid1 pool, it won't increase available space, also not a good idea to add a device with a full filesystem, besides the device you added dropped offline (check cables), but first you need to free up some space on cache, move or delete some data.

[Mesias]

The UI is showing the cache space increased after adding the drive.

 

 

No writes into the new drive so I assume that is just a limitation. Well, I deleted some files and now the cache utilization is below the capacity of the original, smaller cache drive. Can I go ahead and perform a balance to convert the cache to Raid0 while having still some data in the drive?

[trurl]

9 minutes ago, Mesias said:

UI is showing the cache space increased after adding the drive

 

 

[trurl]

12 minutes ago, Mesias said:

Can I go ahead and perform a balance to convert the cache to Raid0 while having still some data in the drive?

Single mode is what you want instead of raid0

 

[trurl]

But it might make more sense to reconsider how you are using cache.

 

Cache should only need to be as large as needed for those things that stay on cache (appdata, domains, system shares) plus enough to cache data for a typical single day of writing to whichever user shares you think need to be cached.

 

If you are needing to write a lot more than typical for some reason, such as the initial data load, then don't use cache.

[trurl]

You have a cache-only share anonymized in diagnostics as d-------s (downloads perhaps?).

 

Why is that a cache-only share?

 

There are probably better ways to do whatever you are trying to do with that.

[Mesias]

Yes, it's a "downloads" pivot folder monitored by other apps. The system was doing just fine with the small SSD, the issue started once I started downloading too many large files. The files needed to stay longer for seeding which tested the limit of the drive. Definitely poor balance from my side, nothing to blame on the software. I now want to increase the cache size to have more capacity. After reading the FAQ in more detail, I'll balance the cache into single mode.

Edited October 9, 2020 by Mesias