How to reconnect Greyhole storage pool after Amahi rebuild

[gboudreau]

Check in /var/log/greyhole.log to see what it's doing.
For me, a fsck of 8TB takes about 2h when it has nothing to repair. Add to that the time it would take to re-create all symlinks, and rewrite all tombstones, I'd guess 4-6h would be about right.

If fsck is done and you still don't see files, post a part of your greyhole.log where the fsck tried to work on a specific file.
Or, just run --fsck --dir /var/hda/files/some_share/some_small_folder and post what appears in greyhole.log when that runs.

[ralfb]

Hi again and many thanks

This is my entire greyhole log file:

Jan 23 03:26:03 7 prerotate: More than 900 tasks queued... Won't queue any more at this time.
Jan 23 03:26:03 6 prerotate: Samba log parsed using --prerotate
Jan 23 03:26:19 2 fsck: Can't insert write task: database is locked
Jan 23 03:26:19 6 postrotate: Samba log read pointer reset using --postrotate

does not seem like much is happening

Also ran as suggested:

Code: Select all

[ralf@padme ~]$ su Password: [root@padme ralf]# run --fsck --dir /var/hda/files/books bash: run: command not found [root@padme ralf]# run --fsck --dir /var/hda/files/books bash: run: command not found [root@padme ralf]#

[ralfb]

It just turned midnight here in Hong Kong and the greyhole log file exploded with activity.

Let's see what happens,

I'm going to bed now.

[gboudreau]

fsck: Can't insert write task: database is locked

That means you're using SQlite, not MySQL. Migrate ASAP.

The actual command I suggested is:

Code: Select all

greyhole --fsck --dir /var/hda/files/books

but you shouldn't need it, is a full fsck is ongoing.

[ralfb]

ok great, migrated to mySQL:

Code: Select all

[root@padme ralf]# /usr/share/greyhole/db_migration-sqlite2mysql.sh Stopping Greyhole service watchdog... rm: cannot remove `/etc/monit.d/greyhole.conf': No such file or directory Reloading monit: Reinitializing monit daemon Done. Stopping Greyhole service... Shutting down Greyhole: [ OK ] Done. Creating MySQL database and user... hda-create-db-and-user: creating greyhole Done. Modifying Greyhole configuration to use MySQL... Done. Migrating data from SQLite to MySQL... Done. Starting Greyhole service... Starting Greyhole ... [ OK ] Done. Starting Greyhole service watchdog... Reloading monit: Reinitializing monit daemon Done. [root@padme ralf]# grep db_engine /etc/greyhole.conf db_engine = mysql [root@padme ralf]#

and restarted greyhole --fsck

Code: Select all

[root@padme ralf]# greyhole --fsck fsck of all shares has been scheduled. It will start after all currently pending tasks have been completed. [root@padme ralf]#

I can now see my shares being populated again so it believe it worked well.

Anything else I could do?


Maybe BALANCE? As I have 3 2 TB drives almost completely full and 1 2TB completely empty.

[gboudreau]

Yes, a --balance would help, well, balance your drives' free space, if you want that.
But it might not be necessary, once Greyhole is done with fsck.
Anyway, it won't hurt either way.

[ralfb]

OK great, thanks.

Started a balance as still one drive empty whilst the others FULL.

Code: Select all

[root@padme ralf]# greyhole --balance A balance has been scheduled. It will start after all currently pending tasks have been completed. This operation will try to even the available space on all directories included in your storage pool. [root@padme ralf]#

[ralfb]

HDDs are nicely balanced now.

Many thanks for your help.