AWS Thinkbox Discussion Forums

balancer not starting nodes

Hi Guys

Trying to get VMX working with AWS, and am not having any luck getting any nodes to pick up tasks. I have managed to follow all of the testing steps up to the point at which I need to start using Balancer.

Screenshot of the balancer log attached. The second last line I’m suspicious of.

Cheers

Fergus
balancerlog.jpg

Hi Fergus,

Thanks for the report. We are looking into it. Can you tell us the exact Deadline build being used when this error occurred?

Hey Fergus,

It would also be really helpful if you could send us a screenshot of the info you’re using to connect to AWS in the Configure Cloud Providers dialog. I think there’s some small thing we’re missing that’s causing the problem.

Thanks,
Eric

7.1.0.21 is the build. I’ll forward some further screenshots when I get to work in a few hours.

Hi Guys

Screenshots of the cloud configuration attached.

On thing that might be relevant. Our instances are hosted in an availability zone ‘b’ on AWS. Apparently there is an ‘a’ and ‘b’ zone in Sydney. This doesn’t seem to be picked up anywhere in the interface. I’m not sure if it needs to be?

Fergus



cloudOptions1_BalancerOptions.jpg

Hi Eric

Now up to 7.1.0.23, and still appear to have the same problem.

As far as we can tell, the subnet and key pair are correct.

Everything it is reporting seems right, (eg. 2 nodes, a job with 16 tasks, the right region, and the right group), but the job just says ‘Queued’, and nothing else happens.

Regards

Fergus
wb_deadline_balancer_150331.jpg

Indeed, the Balancer is not actually launching instances. The “Object reference not set to an instance of an object.” message strikes me as a bug. That shouldn’t happen even if the configuration is wrong. I think we need to set up a remote session for this.

James

I now seem to have it working, but with balancer on a different machine.

Initially I had balancer on the server hosting the database /repository (I couldn’t see anything in the docs to indicate that we couldn’t do this), which may be the issue?

We currently have the database/repository hosted on Amazon EC2, and were trying to avoid spinning up additional servers if we could avoid it. I can find somewhere else to run balancer that will always be online, but it would be preferable if we could keep all the components in Amazon.

Regards

Fergus

Hi Fergus,

Glad to hear you got it going. Balancer is fairly lightweight and should be able to run on the same instance as the one hosting the DB/Repository (assuming the network interface has sufficient bandwidth). I can only speculate that there was some sort of installation or configuration issue. If you decide you want to continue to investigate, just let me know.

Privacy | Site terms | Cookie preferences