Slow Image Deployment
-
Hi,
I’m trying to deploy an image to a computer on my LAN (the main fog server is also on my LAN), but it’s choosing to pull the image data from a storage node in a remote office through a slow connection instead of the main fog server on my LAN. Why?
The latency to the remote storage nodes varies from 30-60 ms, but the latency on the LAN is <1 ms. It makes no sense as to why it’s pulling the image from a remote storage node.
FOG Main Server OS: Ubuntu 24.04
FOG Version: 1.5.10.1660 -
@ProfessorFow If you create storage nodes, and this is the worry, separate them by Storage Groups.
My guess is all your storage nodes, right now, Remote/local/etc… are all part of the same storage group.
It doesn’t know your latency situations or that any specific node is “Remote” vs. local.
All it knows is “I have x nubmer of nodes in my group, I’m going to use the one with most resources available to run to perform the task at hand.” if no nodes are busy at the time, it’s whatever is the first to return.
If you have “local” nodes, keep them in a group, and if you have remote nodes, keep those remote nodes localized by their own storage group.
Storage groups can be any method you choose, though I’d recommend keeping subnets as different storage groups.
-
@Tom-Elliott Lol they are indeed all under the default group storage group. I was hoping to have the primary FOG server replicate to the remote storage nodes. If I separate them into their groups, will I no longer have automatic replication from my primary fog server to the remote storage nodes? How would I get around that?
-
@ProfessorFow There is the ability to replicate images across groups. You can set which “groups” images need to replate to by associating images to multiple storage groups. You can even define which storage group is the primary.
