I have a Hybrid Core with 3 HDD's:
200 Gb IDE drive with the OS on it and a small amount of media
450 GB SATA drive with 200 GB of Video on it
1 TB SATA drive with 770 Gb of Video and Audio on it
with the core alone working, all media is visible on the orbiters.
If I boot up any one of my two MD's (a Dell D620 with Gigabit Broadcom NIC or an Athlon XP 2600 desktop machine with 100 Mbit NIC) and then access some of the media, after an amount of time the list of media changes and only a few files are visible, some on the 200 GB drive and one or two on the 1TB machine (Thom's 'using media in LinuxMCE' weirdly and an REM album). After this I have to reboot to get it back. before the reboot I can go into the KDE desktop and use dolphin and access the media from there on the core.
Any ideas why this is happening and what I can do to get around it?
Sounds like this is the long standing problem with the StorageRadar scripts... it typically affects systems with multiple data drives added to the Core and or cifs shares.
We use a pretty ugly temporary fix for this currently that forces the drives to remain mounted using a screen session;
open an ssh session on your Core and start a screen session... then cd to /home/public/data/videos/<symlink to 1st of your data drives>
so this goes something like this;
- ssh into your Core
Use the screen command;
screen -S keep_drives_mounted<return>
**The string after the '-S' switch above can be anything you desire... its just to make identifying the screen session easier ;-)
cd /home/public/data/videos/<symlink to 1st of your data drives> <return>
Now open a new window with "Ctrl-A" "c"; ***That is hold Ctrl down and press 'a' once...then with no other keys depressed...press 'c' once ***
and 'cd' as above but this time into your next drive.
Now open a new window with "Ctrl-A" "c"
and 'cd' as above but this time into your last drive.
Now you can switch to each window to check they are all there with "Ctrl-A "n" which will cycle through each screen window each time you do it.
Now you need to detach from the window using "Ctrl-A" "d". This will drop you back into your shell. All screen windows you created are still there and you can re-attach to them later at anytime by doing the following to list all your screen sessions;
screen -ls <return>
You should see something like below;
demo@DCErouter:~$ screen -ls
There is a screen on:
13022.keep_drives_mounted (15/03/09 12:44:25) (Detached)
1 Socket in /var/run/screen/S-tracy.
demo@DCErouter:~$
Now you can see the screen session called '13022.keep_drives_mounted' listed as 'detached' ... so we use the following command to re-attach to your screen session;
screen -r 13022.keep_drives_mounted<return>
Normally once you have set this screen session up you should never need to go back to it. If you reboot your Core then all of the above screen sessions will be lost and you will have to re-do the steps above again... or create a script that runs during boot up to do it for you.
The above procedure is a 'hack' and hopefully this problem will be resolved properly in the StorageRadar scripts for 810.
All the best
Andrew