On Sat, Sep 9, 2017 at 8:25 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com> wrote:
Hello,
it is not clear to me the exact workflow in case of testing what in object.

For sure first step is to install base node from iso.
NGN has no more read only filesystem.
Does this mean that I have to set up eg multipath for FC LUNs (engine and data domain) before running the engine setup via cockpit?
Or when I specify FC in type of storage domain it would ask and setup automatically multipath for me?

No, it doesn't: HBA and mutipath are supposed to be directly configured before running hosted-engine-setup.
It's exactly the same as for vlan and bonding.
 
Are the ovirt repos already setup when I have installed from the iso or do I have to install ovirt-release-xx rpm?
In another enviroment with block based storage domains (iSCSI) and an external engine and full CentOS hypervisors I customized vdsm creating a file /etc/vdsm/vdsm.conf.d named 50_thin_block_extension_rules.conf using
volume_utilization_percent = 25
volume_utilization_chunk_mb = 4096

Can I make these sort of customization in NGN without any persist command, because the file system is read write now or should I manage from cockpit or other tools instead?

I'm exploring situations where NGN could be suitable better than full CentOS OS for hypervisors and I would like to know clearly advantages and limits.

Thanks in advance

Gianluca

_______________________________________________
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users