Image Computers with logical batch processing?



  • Hi,

    we have around 150 computers i like to image in a fixed time schedule, i can do this in the following time windows:

    [INDENT=1]- monday-friday from 7:30 pm - 07:00 am[/INDENT]
    [INDENT=1]- saturday-monday from sa 4: pm - mo 07:00 am[/INDENT]

    I think it’s enough to have a weekly image for each computer.

    I think the average imaging time for 1 computer that is imaging exlusivly with around 20-50gb data will take 10-25min to image.

    Counted in hours it’s 62,5 hours these are 3750 minutes divided by in worse case 25 minutes = lol 150 pffff :D

    But…

    Howto time the schedule plan optimal? Afai can see there is no intelligence when scheduling imaging task like only 3 clients at same time, check when next client is finished to immediately start the next one and so on.

    Am i wrong or how do you manage such cases?

    Greetz X23


  • Moderator

    There is no mass upload built into FOG. It’s meant more for a mass deploy. That being said, it’s open source and can be modified to do whatever you want. Sorry.



  • @chad-bisd, post: 12067, member: 18 said:

    Each storage node has a maximum number of slots it will image concurrently. Once that is reached, it queues the rest.
    means if i have 4 scheduled images that may overlap fog let 3 computers run and if the first is finished it starts the 4 one also if the schedule time is overdue?

    @chad-bisd, post: 12067, member: 18 said:

    The clients check in every 10 seconds to see if a slot has opened up and if they are next in line to grab that slot

    Sounds like you talk about the fog windows client? will it reboot a computer if an image is scheduled so it can boot over pxe to make the image?

    @chad-bisd, post: 12067, member: 18 said:

    I think you just need to select all the machines you want to image, throw them into a group, and tell the group to deploy the images. It should queue them and process more as they finish.

    I don’t want to mass deploy i want to mass create images with an intelligent mechanism.

    Greetz X23


  • Moderator

    Each storage node has a maximum number of slots it will image concurrently. Once that is reached, it queues the rest. The clients check in every 10 seconds to see if a slot has opened up and if they are next in line to grab that slot.

    I think you just need to select all the machines you want to image, throw them into a group, and tell the group to deploy the images. It should queue them and process more as they finish.

    If all machines are the same image, consider multitasking. If not, set the storage node max # of clients to 3 so that only 3 are ever imaging at the same time.


Log in to reply
 

367
Online

38721
Users

10548
Topics

99849
Posts

Looks like your connection to FOG Project was lost, please wait while we try to reconnect.