Pre-flight checks out. AWC says all is good their side but the instance never launches after I submit. Just checking to see if anyone else is having this problem? Or if I need to buy a Mantra token although I think Indie supports one Hqueue job.
Thanks, Kevin

I've been having some issues with cloud rendering via Indie as well. This is a copy / paste from a thread I made on the OD Force forums:

I've sent a couple jobs out but seem to have hit another snag. After I hit render I get the normal preflight / password dialogs, my browser opens to the Cloud Submit page, but after a while I get a dialog box with the message “The Houdini Cloud Submit daemon has terminated prematurely.” There is a direction to a log file - this seems to be the most relevant passage:

ERROR: AuthFailure: Unable to launch image ami-73607f1a as the product code 5D6864EE is not supported for the specified region, instance type and operating system combination. Please contact the vendor for more information.

Any similarities to your troubles, Kevin?

My first thought was also that it could be an issue with the version of Houdini on the cloud machines not being able to process files from Indie, so hopefully someone can shed some light on that for us.

Hi John,
It might be. Houdini's EC2 browser page warning was looking for the Houdini version. Then in Jobs, the last dialogue before it stalls out is that it has initialized 13.359 (the last production build before .509 I think) but I'm running Indie .509 so it may be a versioning problem.

I submitted a support ticket which they been jumping on like a trampoline. Last iteration they asked me for my cloudsubmit.log which (on a Mac) is both hidden from view and buried. If you sudo find, etc in your terminal, you can find it no problem though. I sent mine in today and will let you know what they tell me.

Kevin

Hmm… seems you're getting further than I am. I don't see anything about any version of Houdini being initialized.

In any event, I think I'll submit a ticket as well and see where that leads - I'll also keep you updated with any progress on that front.

Support just replied

“Our developers agree that this could be a versioning issue. We will update the cloud version this week.

For future reference, if a job fails on the cloud, you can find out why by clicking on the View HQueue Jobs link on the cloud web page. Then keep clicking/diving into the failed Child Jobs until you reach the lowest job. The Command Output text area should show the error that caused the failure. You can also download the output.”

Kevin

Good to hear. Perhaps the update to the cloud version coincides with the release of the new production build today? I've just installed it and will try to fire off a render later tonight.

I'm having a similar issue I can only seem to launch a single machine type through amazon which is the 20 ec2 Compute units with 7 gb's of ram. I hope they fix this soon. I should note I'm using a fully licensed version, but am still having a similar issue.

TRex, Did you try fewer machines? I think I read somewhere that Amazon has a 20 machine instance limit.

Oddly, I think SideFX has something about the 28 machine option being the optimal cost/performance choice, or I could be entirely off base here.

Kevin

TRex92
I'm having a similar issue I can only seem to launch a single machine type through amazon which is the 20 ec2 Compute units with 7 gb's of ram. I hope they fix this soon. I should note I'm using a fully licensed version, but am still having a similar issue.

This seems to have done the trick… kind of. If I select that machine type my jobs will go through, only to fail with the same message that Kevin has been getting this whole time. TRex, are you able to get finished jobs from that machine or are yours also failing?

Kevin, the Compute Units thing is a little confusing. Basically it's Amazon's metric for the processing power in a single machine, so you can pick a machine with however many Compute Units but it's still only a single instance unless you specify that you want more in the HQueue Render properties.

Thanks John, good to know. Installing the latest prod build now…I'll give the cloud another shot later.

Now working but per support these configurations are still problematic

- 7 EC2 Compute Units
- 14 EC2 Compute Units
- 28 EC2 Compute Units
- 55 EC2 Compute Units
- 108 EC2 Compute Units

The 20 instance worked fine in my first test run.
kevin