AWS Thinkbox Discussion Forums

Rendering with Clarisse in the cloud

Hi,

We are playing with the latest deadline at the moment with assets server and cloud rendering with Clarrise 4.

Does anyone currently do cloud rendering with Clarisse?

We are having issues with the asset server, getting the aws machine to pull the relevent assets through.

It looks like it connects ok and I have opened the firewall for port 4000 and also a static port forward but says it can’t find the asset.

I’m wondering if anyone has got around this issue and is currently using the cloud machines for slaves.

On a side note, when will you officially support Clarisse 4 please?

Many thanks
Mark

Hi Mark,

Is the error you are referring to in the Asset Server log?
If yes, can you please email support AT thinkboxsoftware DOT com with a copy of the relevant log so the support team can see what you are seeing?

Can you also please specify the OS type and version you are running on-prem and on AWS, and the exact version of Deadline involved?

Clarisse is not officially supported by AWS Portal out of the box, but it might work as long as the Path Mapping can be applied to the source scene to trigger the asset transfer. So if you are getting errors in the Asset Server log, it would imply the Asset system is being triggered correctly by the render nodes, but the actual service fails to deal with the transfer. If the error was in the Slave Log, then it would be a different issue…

Hi Mark,

we rendered a show with Clarisse on AWS about year ago, it think it was 3.6 back at that time. The AWS portal was quite new and I struggled a bit with the documentation which was lacking a lot of details back then.

The initial setup was definitely a pain, but more on the licensing side of things, I can’t recall having any bigger issues with the asset transfer. Opening firewall ports and/or setting up port forwarding shouldn’t be necessary, since the ssh tunnel used for file transfer is opened by the machine running the awsportallink. It’s quite hard to guess where your problem lies with these limited information, but feel free to ask, if you have specific questions.

Best, Til

Hi All,

Thanks for the replies.

I’ve found I had a few errors in the portal link log so now fixed them.

I’m at the stage where I have got it to find everything I thik, but it says the output doesn’t exist that we create from a pre-script. It now converts the path on the render node:

2019-09-06 13:04:09: 0: STDOUT: 00:00:00 87/87MB ERROR: Objects with invalid output filepaths:
2019-09-06 13:04:09: 0: STDOUT: Object: ‘project://scene/image’: Target folder ‘E:/lsalinuxtest.locksmithanimation.comrepository478e3c54c06723d120768849e4612074/output/brian.silva/render_testing_B.v5/image’ does not exist.
2019-09-06 13:04:09: 0: STDOUT: Object: ‘project://scene/image_witness’: Target folder ‘E:/lsalinuxtest.locksmithanimation.comrepository478e3c54c06723d120768849e4612074/output/brian.silva/render_testing_B.v4’ does not exist.
2019-09-06 13:04:09: 0: STDOUT:
2019-09-06 13:04:09: 0: STDOUT: Please check that:
2019-09-06 13:04:09: 0: STDOUT: - output paths exist and have write permissions.
2019-09-06 13:04:09: 0: STDOUT: - ‘save_as’ and ‘save_deep_data_as’ attributes of Images / 3D Layers are correctly set.
2019-09-06 13:04:09: 0: INFO: Sending EndTaskRequest to S3BackedCacheClient.

So i can see it’s converted the mapping but I don’t think that the pre-script has created the paths on the render node.

It fails to run the pre-script if we try it on a render node, it runs fine on the local render nodes that can see the network map.

Is there a way I can get onto the render node via remote desktop, my pem file doesn’t seem to work.

Thanks
Mark

OK, we think it’s not creating the output folder that is created in the pre script, so just trying that.

Is there a way to use my in house cnode licenses for the slave machines, I have added the license server and port number but has not worked.

Thanks
Mark

Privacy | Site terms | Cookie preferences