Page timeout when creating multicast task

[dolf]

FOG version 8595
Ubuntu 14.04
Accessing FOG web interface from Google Chrome 51.0.2704.106 m

When creating a multicast task using the web interface, the next page only loads after all of the PCs have turned on (they so so sequentially - it’s quite a beautiful thing to see). During that time, the browser shows “loading”. Now when I do this with many (50+) PCs, either the browser or PHP (unsure) becomes impatient at some point and terminates the script prematurely. The result is:

• the multicast task is created, and waits for 50 hosts to join
• n < 50 tasks are created for hosts to join the multicast task
• m < n PCs have actually turned on (because of WOL)

Maybe this is due to PHP script timeout? Maybe it is due to the browser deciding that the server must be dead if it doesn’t respond in time? In either case, it’s not scalable. I would like to task ALL my hosts at once.

It would be better if there were a background process to create the tasks, so that the next page can load while that continues.

[Tom Elliott]

Please try updating. I’ve added code to much more quickly create Multicast Tasks (and more accurately I may add). I’ve also added code that, for non-multicast tasks, will run in the background while you still can move around the pages. Hopefully this helps in all respects. This, for now, does mean multicast from groups will not generate the requisite snapin taskings, but it’s only because I need to think of the logic to enable it. It should work hopefully soon.

[dolf]

Thanks! It flies now.

[Tom Elliott]

Glad I could help.

Just an update, snapins should now be deployed for Multicast Tasks.

I will work to update the other tasking types in a similar fashion.

Thank you,

[Tom Elliott]

Now all group tasks are created doing a batch insert method. This should prove extremely efficient.

The only part that isn’t handled is WOL once the tasking is created, however this shouldn’t pose much an issue as the FOG Scheduler service (normally cycled once every minute) will check for any tasks awaiting checking and send a WOL packet to those hosts. It may take a little longer, but the same result is achieved I think.

[dolf]

Great! I was about to say that the same goes for group unicasts. WOL by the FOG Scheduler service should be fine.

[Wayne Workman]

@Tom-Elliott said in Page timeout when creating multicast task:

This, for now, does mean multicast from groups will not generate the requisite snapin taskings, but it’s only because I need to think of the logic to enable it. It should work hopefully soon.

I’ve been learning about triggers in mysql. It’s really super efficient, you could do all the snapin stuff from deployment with triggers.