• Good Afternoon,

    S0, I’m preparing to rebuild FOG.

    I have 1 VMware host with 2 datastors and I would like to utilize them both. Each datastore will be vmdk on the vm.

    I will install Ubuntu and FOG on datastore 1.
    I would like to store some of the images on the second datastore.

    How can I set this?

    In addition, It there a way to see (from FOG) how much space each image takes on the disk (after compression)?

    General question:
    How much free space would you leave as overhead from the datastore? (for snapshots, consolidation etc…)



  • Senior Developer

    @Tom You reported the same issue here. I tried to reproduce but I can’t. Please check if /images/<image-name> is still present after deleting the image definition?!?

  • @george1421 Thanks George. The issue with the graphics is that it will not update.
    Earlier, I tried to upload an image and it broke in the process.
    The indicator showed 15 GB was used. I deleted the image and checked the box to delete the file. Refreshed few times and waited about 15 minutes, but the indicator still showed 15 GB are used.

  • Moderator

    When you capture an image the gui interface is updated with the size of the capture. This isn’t real time but at the time of image capture.

    If you are wanting to see about total space consumed, on the dashboard there are graphics that are real time that show free and a used space on a per storage node bases (The main fog server is also a storage node)

  • @george1421 Thank you for your response. After reading your post, I decided to simplify things, hence I rebuilt the host with 1 datastore.

    Is there a way, in FOG console, to see the image size on the disk?

    Thanks for all you do.


  • Moderator

    You can kind of do what you want with a few restrictions.

    Actually I (personally) do recommend that you put the FOG server root partition on one vmdk and the target images on a separate vmdk file. This helps from a performance as well as being able to expand the /images data store. AFAIK you can’t split vmdk files across data stores, so each vmdk file must exist on a single datastore.

    What you can do is this.
    1 vmdk for the OS and FOG
    1 vmdk for /images
    and if you use snapins another vmdk for that one. You can scatter the vmdks across your datastores (realize this isn’t best practice). Its best practice to place all vmdk files in a single folder structure. Just because its best practice doesn’t mean you must.

    Here is a KB to my recommendation
    1 vmdk for Linux OS and FOG
    1 vmdk file /opt (which will contain /images and /opt/fog/snapins)