Amahi Forums

I've just installed Amahi 11 and I've got a curious issue with shares and the Greyhole pool. I was previously running Amahi 10 without issues and got a wild hair to upgrade to 11. I have an HP microserver with a 1TB system drive and 3 1TB data drives in a greyhole pool. I followed the guidance to install Fedora 27, Amahi, and then reconnect the data drives. I had an issue with partitioning (see bug #2347) but was able to work around and get it installed. The issue I'm now having is that I can't access any shares that are in the greyhole pool from a client on the network. If I remove a share from the pool using the Drive Pooling tab on the dashboard I can then see it and map it from a client. If I check the box and add it back to the pool it immediately becomes inaccessible again. I'm assuming it's some sort of permission issue but I'm not sure where. All files at /var/hda/files are owned by the admin user created during installation. Incidentally, the mount_shares_locally script can't access any files in the pool either so all /mnt/samba/xxx directories are empty. Any guidance as to where to look for the problem is appreciated.

https://paste.fedoraproject.org/paste/z ... RZxaLXwJzg

You might try a restart.

Sometimes Greyhole and Samba need to be restarted. You can go that route or just do a reboot.

I have not fully upgraded to Amahi 11 yet, so don't know if this will be an issue for me.

I presume this was a new install, not an upgrade.

Thanks. I've rebooted it several times thinking that might help. Yes, this is a new install. I used the upgrade path when I went to Amahi 10 and it worked flawlessly. I first tried it here with 11 but it didn't work too well so I went with a clean install.

Getting a lot of cifs errors. For example:

CIFS VFS: BAD_NETWORK_NAME: \\127.0.0.1\Pictures
CIFS VFS: cifs_mount failed w/return code = -2

This happens for each share in the pool.

There may be an issue with the mount_shares_locally script for Greyhole causing your shares issue.

See the Greyhole wiki page.

You will need to submit a bug report to the Greyhole author potentially. Might be a compatibility problem with Fedora 27.

Just noticed I have the same error on my HDA.

Interesting part is I am still using Amahi 10 on Fedora 25.

I disabled mount_shares_locally and will see if that changes anything. My shares worked fine with and without it enabled.

I just reinstalled 11 and checked to see if the shares were accessible before installing Greyhole and reestablishing the share pool. Although empty, each share was accessible from a Windows 10 client. Before doing anything else I installed the Greyhole GUI app and added the data drives back in following the guidance docs provided for Amahi 11. Then, if a share was added to the pool it became immediately inaccessible. If removed (unchecked in the app) it became immediately visible. I didn't run a fsck. I think there's a bug somewhere either in 11 or in Greyhole.

You may need to run a fcsk first on a single share.

Try that and see if anything changes.

I will try to test more as well.

I was reading through this page https://github.com/gboudreau/Greyhole/w ... ole-Shares at the Greyhole wiki which describes my situation exactly. When I list the /usr/lib64/greyhole directory I get

[root@hda greyhole]# ls -la
total 260
drwxr-xr-x 2 root root 4096 Oct 5 17:20 .
dr-xr-xr-x. 55 root root 32768 Oct 5 17:04 ..
-rw-r--r-- 1 root root 40126 Jun 21 2017 greyhole-samba40.so
-rw-r--r-- 1 root root 40594 Jun 21 2017 greyhole-samba41.so
-rw-r--r-- 1 root root 44306 Jun 21 2017 greyhole-samba43.so
-rw-r--r-- 1 root root 44106 Jun 21 2017 greyhole-samba44.so
-rw-r--r-- 1 root root 53064 Dec 28 2017 greyhole-samba45.so
lrwxrwxrwx 1 root root 39 Oct 5 17:20 greyhole-samba47.so -> /usr/lib64/greyhole/greyhole-samba47.so

Which implies to me that the greyhole-samba47.so file isn't there, only a symlink. Correct? Am I reading this correctly?

Could be a problem.

Try this as root: Then restart Greyhole: