Jump to content
Updated Privacy Statement
  • 0

Delivery Group Load Balancing - does it avoid power state unknown machines if it can? (7.13)

Richard Hall1709161984




We're trying to work around a problem on a particular virtual infrastructure, but we're seeing some odd behaviour. I wonder if anyone knows if this is expected or not?


We have a Delivery Group which contains 10 machines from Machine Catalog 'A' (the entire catalog) and 1 machine from Machine Catalog 'B' (also the entire catalog).

Machine Catalog A is Server OS (Virtual), PVS, Random Allocation

Machine Catalog B is Server OS, PVS, Random Allocation



All of the machines in Machine Catalog A are online but appear in Studio as 'Power State': Unknown, 'Registration State': Registered

The single machine in Machine Catalog B is online and appears as  'Power State':Unmanaged, 'Registration State': Registered


If we either remove or maintenance mode the single machine from Catalog B, the sessions are split evenly across the 10 servers in Catalog A - Great.

However,  once we add the machine from Catalog B back into the mix, it appears that all sessions are being sent to it, despite the 10 other machines being sat there with no sessions at all! 


As soon as we get rid of that Catalog B machine, things go back to normal and spread across the A machines.



Is Studio doing something in the background where it is favouring the 'physical' machine? Or is it just avoiding the unknown power state machines (even though they are registered OK) unless it has no other choice?


In reality, the B machine is identical to the A machines (has come from the same PVS image etc.) but as you might be able to tell from the unknown power states we're having some issues with the Studio <> VI link which we're trying to work around.


Thanks for any clarity anybody can provide,


Link to comment

1 answer to this question

Recommended Posts

  • 0

As an update to this, we experimented with moving another machine from Machine Catalog A to B and then (carefully) added them into the Delivery Group. As before, all sessions were immediately taken by the Catalog B machines, but at least they were balanced between the two machines from that Catalog this time.


We then very quickly moved all other machines into Catalog B and re-added them to the delivery group, so that things were balanced amongst all 11 as we were originally aiming for.


On to the next problem...

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Create New...