Fog Host Management gives Blank Screen after SVN up to 6067
-
Server
- FOG Version: 1.3.5
- OS: Fedora 22
SVN Revision: 6067
Client
- Service Version:
- OS:
Description
Updated SVN to 6067 (from 5955, I believe) today and everything looks fine, but the Host Management screen is nothing but a white screen.
I can see the hosts if they are attached to a group…and can even see all the hosts by clicking ‘Membership’ of the group and then selecting the option to ‘see hosts not within this group.’
Any help will be greatly appreciated with the ability to see hosts on management screen.
Thanks. -
How many hosts do you have?
Probably out of memory I’m guessing.
You can set this higher from FOG Configuration Page->FOG Settings->General Settings->FOG_MEMORY_LIMIT
-
It is currently on 128. Do you have a higher recommendation or does it matter?
-
@ftgjason I typically follow the 128 rule.
128 -> 256 -> 384 -> 512 etc…
-
@Tom-Elliott I have 1653 hosts found. It finally loaded at 256. Is there any hardware restraint for this setting?
-
@ftgjason The memory your server has.
1653 is not that many hosts, but I can definitely see there being more memory use than anticipated for things. I’m trying to figure out a nicer way to handle information gathering but it probably won’t show up for a while.
Just try to ensure you leave enough memory for any other applications (NFS, FTP, TFTP, etc…) but it’s also why I follow the 128 rule. 1024 is what I have my systems on but I typically am testing with upwards of 6 - 10k hosts (so I need a much larger bucket.)
-
@Tom-Elliott Thanks for your help.
-
@Tom-Elliott Also noticing that in the Image Management, there are 0.00iB on all my images sizes: ON SERVER. Made me think that I had lost all my images, but we were able to fog a machine successfully.
-
Those fields are “informative” elements. They are not meaning you actually have, or not, images. Those are updated by a service called
FOGImageSize
now, in the past it was handled by FTP requests (which slowed things down significantly). Either permissions aren’t correct on the locations, or it just hasn’t had enough time to update the information as needed. -
@Tom-Elliott Great. Will wait to see what happens on the size listed. One more question. Do you have any idea why our PXE boot has gotten slow. It does the main PXE boot and then puts nearly 2 rows of . . . . . before the machine will begin going to windows or doing a job from FOG.
-
@ftgjason Are you using pxelinux.0 as your boot file or are you using undionly.kpxe, ipxe.pxe, ipxe.efi? (Just trying to get a good understanding).
-
@Tom-Elliott undionly.kkpxe is 067 on windows DHCP server
-
Then I can only speculate it’s transferring the file EXTREMELY slowly. Just enough to keep it alive, but slow enough to be annoying.
Maybe a restart to the fog server might help?
-
@Tom-Elliott will try a reboot. If you think another boot file would be better, please let me know.
Thanks man. You are extremely helpful. Have a great day!