iSCSI configuration questions

[jayst136]

i'm deploying a 3-node 8.0.13 cluster connected to iSCSI storage. GFS2 storage.

I've read the deployment guide. 
Why is the iSCSI configuration documented at the bare minimum compared what i saw in the past?
I've now configured the nodes with two iSCSI interfaces in two subnets. I've not touched ANY iscsid configuration at this point.
All good, but then i went to the VME Manager and just added two target ip addresses of my array under storage and away we went. Multipath devices were present and sessions were made. This is different from all the steps we had to do in the past regarding iface, session logins etc etc. 

i'm now a bit lost in what to expect from VME software regarding iSCSI setup and what's left to worry about before VME goes at it.
Do we still have to take care of iscsid configuration regarding session login , iface binding etc? or is VME Manager handling (almost) all?

[CalvinZito]

Hi @jayst136 - via one of my technical experts:

All that typically needs to be done is adding the iSCSI targets, then selecting the discovered devices when configuring a new datastore. Those pieces that had been manual in the very beginning are now handled by the UI.

[jayst136]

Hi Calvin,

Ok, that's good to know.

However, this default could quickly be less than optimal. HPE started with incuding the OS in he deployment model, things like this should come optimized or instructed to optimize afterwards. Just like we have documented the process of configuring a mii-interval for network bonds to make a bond failover in de first place (which is still very strange imho, but it's an ubuntu thing). 
In regards to iSCSI, for example, VME does not seem to do anything with the mutipath configuration although it's documented to be adjusted (for Alletra only?). 
All in all, this could use some work, at least in the documentation in my opinion.