IP Allocation on large deployments fail

[jakubjs]

Hi

I have a blueprint that deploys a number of linked clones. This was working fine when the networks where set to DHCP. I have changed it to Static allocation and now when I deploy more the 35 vms it sits for an hour and then says time out. If I request 32 it take a couple of minutes to do its Allocation process and then happily builds all the vms. Is there some parameter that needs to be changed and where would I be able to see what its trying to do?

cheers,

Jakub

[daphnissov]

Is vRO doing the allocations from an IPAM framework or custom workflow? If so, the vRO app-server might be out of memory. You might want to check the logs to see if this is the case and increase the Java heap size appropriately.

[jakubjs]

Hi

Thanks for your reply. We are using an distributed deployment of vra 7.2. So we have 2 of each vra, vro and iaas servers. The ip allocations are done via the default framework. Which server would be the one that is handling the actual allocation. I'm assuming this would the the windows IAAS server as I can see that the MSSQL server holds the ip allocations. Is this correct? Maybe they are underspected.

cheers,

Jakub

[daphnissov]

How many DEM workers do you have? By default, they can process 30 workflows concurrently, and right over that number. If you're going to be executing that many simultaneous builds, you should have at least two and possibly three. Next time a deployment runs, you can look at the DEM resources for utilization and in the DEM worker tab under Infrastructure to see how long workflows are taking to execute and if they're hanging.