Strange behavior for images location
-
Hi,
i’ve got a fog 0.32 with 2 hard drives. [B]The second[/B] is used for my images.
In the fog interface,
i created 2 storage grooups (disque1 and disque2)
The image path in the image management is like /media/disque2/name of the imageI created 2 storages nodes partage1 (link with disque1, image path =/media/disque1) & partage2 (link with disque2, image path=/media/disque2)
I created on group with 2 computers. I associate an image located on the second disk (i tried several diffents images).
If i download in unicast, no problem.
If i try a multicast download, the clients boots, then waits in the process of downloading the image.
In the fog viewer, i have : test multicast failed to execute, image file:/media/[B]disque1[/B]/name of the image not found!I read the fog wiki, this forum, but i didn’t see any solution. Any idea ?
Thanks in advance -
a little up
-
Is there a reason why you still use FOG 0.32? I am still new to this project and don’t know much about the old versions but I suppose this could be a bug in 0.32… and it won’t be fixed as we are heavily working on newer versions.
Maybe you want to setup a new test server with FOG 1.2.0 or even developers version and see if the problem is still there. Please let us know and I am sure we’ll find a way to fix it!
-
Based on the “test” messag eyou have there, does the file actually exist in /media/disque1/nameOfImageHere?
The other thing, this is a customized setup I’m guessing, as you’re attempting to use /media as the mountpoint, and disque1 or disque2 based on the file?
I believe the “bug” as Uncle Frank describes with 0.32 was that it always expected the nfs to be at /images. Due to this, you would need to modify the exports file as this did not occur naturally on the system when making such a change.
Edit the exports file and create the needed entries to your relative mount points:
Most likely entries like:
[code]/media/disque1/images *(ro,sync,no_wdelay,insecure_locks,no_root_squash,insecure)
/media/disque1/images/dev *(rw,sync,no_wdelay,no_root_squash,insecure
/media/disque2/images *(ro,sync,no_wdelay,insecure_locks,no_root_squash,insecure)
/media/disque2/images/dev *(rw,sync,no_wdelay,no_root_squash,insecure[/code]Once those entries are made, you’ll need to refresh the nfs most likely with
[code]sudo exportfs -a[/code]
As these are custom entries, I’d also make sure to first run:
[code]sudo touch /media/disque1/images/.mntcheck /media/disque1/images/dev/.mntcheck
sudo touch /media/disque2/images/.mntcheck /media/disque2/images/dev/.mntcheck
sudo chmod -R 777 /media/disque1/images
sudo chmod -R 777 /media/disque2/images[/code] -
First, i apologize because my fog server wasn’t in 0.32 but in 1877 (version 1.1.0 ?)
I upgraded after the uncle franck message (via [url]http://www.fogproject.org/wiki/index.php/Upgrade_to_trunk[/url]
with the svn method).
No error during the upgrade.
I did what explained Tom elliott in the third message (except i did it only for the second disk, i deleted all informations for the first disk).Problem, in the interface, i coud not have access to the Report Management (blanck screeen with only write version:3390)
In apache2/error.log i saw
[error] [client 10.144.20.2] PHP Fatal error: Call to undefined method FOGURLRequests::resolveHostname() in /var/www/fog/lib/fog/FOGURLRequests.class.php on line 12, referer: [url]http://127.0.0.1/fog/management/index.php?node=home[/url]
Finally, i deleted all storage nodes in Storage Management -> access to Report Management OK
But impossible to create a new Image (Blanck screen with write : add new image definition)
If i add a storage node after that -> No access to report management
But it’s possible to create a new image
Is there a way to fix it ? -
You can upgrade again and that issue should go away.
-
[quote=“Tom Elliott, post: 43677, member: 7271”]You can upgrade again and that issue should go away.[/quote]
thanks a lot !
After the new upgrade, the interfaces bugs disappears.
When i launch a multicast download, in the fog log i can see the right image location.
But I’m still block after the wol, the clients stop at the image of partclone.
I suspect a network issue .
I tried the procedure udp send and udp receiver for 1 client.
[url]http://www.fogproject.org/wiki/index.php/Multicasting#Testing_1_Client[/url]udp-send : ok
udp-receiver : no response arrive to the server.
I will search and let you know what i found.
