Regarding Gluster question. The volumes would be provisioned with LVM on the same block device. I believe 100Gb is recommended for the engine volume. The other volumes such as data would be created on another logical volume and you can use up the rest of the available space there. Ex. 100gb engine, 500Gb data and 400Gb vmstore. 

Data domains are basically the same now, in the past there used to be different domain types such as ISO domains which are deprecated. You don't really need any more than engine volume and data volume.  You could have a volume for storing ISOs if you wanted to. You could have a separate volume for OS disks and another volume for data disks which would give you more flexibility for backups (so that you could backup data disks but not OS for example). 

On Fri, May 29, 2020 at 10:29 AM Jiří Sléžka <jiri.slezka@slu.cz> wrote:
Hello,

I am just curious if basic gluster HCI layout which is suggested in
cockpit has some deeper meaning.

There are suggested 3 volumes

* engine - it is clear, it is the volume where engine vm is running.
When this vm is 51GB big how small could this volume be? I have 1TB SSD
storage and I would like utilize it as much as possible. Could I create
this volume as small as this vm is? Is it safe for example for future
upgrades?

* vmstore - it make sense it is a space for all other vms running in
oVirt. Right?

* data - which purpose has this volume? other data like for example
ISOs? Direct disks?

Another infra question... or maybe request for comment

I have small amount of public ipv4 addresses in my housing (but I have
own switches there so I can create vlans and separate internal traffic).
I can access only these public ipv4 addresses directly. I would like to
conserve these addressess as much as possible so what is the best
approach in your opinion?

* Install all hosts and HE with management network on private addressess

  * have small router (hw appliance with for example LEDE) which will
utilize one ipv4 address and will do NAT and vpn for accessing my
internals vlans.
    + looks like simple approach to me
    - single point of failure in this router (not really - just in case
oVirt is badly broken and I need to access internal vlans to recover it)

  * have this router as virtual appliance inside oVirt (something like
pfSense for example)
    + no need hw router
    + not sure but I could probably configure vrrp redundancy
    - still single point of failure like in first case

  * any other approach? Could ovn help here somehow?

* Install all hosts and HE with public addresses :-)
  + access to all hosts directly
  - 3 node HCI cluster uses 4 public ip addressess

Thanks for your opinions

Cheers,

Jiri

_______________________________________________
Users mailing list -- users@ovirt.org
To unsubscribe send an email to users-leave@ovirt.org
Privacy Statement: https://www.ovirt.org/privacy-policy.html
oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/
List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/LIFQQHTFVTS6KICR5MTRPGO5CH7QDLK7/