Since updating Deadline to 10.0.20.2 I can’t connect to the Aws instances to see what they are doing.
They are able to send a log back once a task is complete, but it’s a big unknown what they are doing inbetween starting the render and finishing!
This is all I am seeing from the slave:
Connecting to IP-0A80321A…
Success: Connected to Pulse, attempting to redirect command to target…
timed out while waiting for data
Remote command indirection is enabled, please ensure that your Primary Pulse is running and reachable.
Any ideas? Got a pulse running locally (and nothing has changed since 10.0.18.1 on our end.
I’ve seen an issue where the Monitor used the Primary Pulse instead of the one on the Gateway. We should send commands to the Pulse based on what region it and the target machine are in… We fixed a bug in 10.0.20 for that one, and the code changed was in the Monitor.
@Berkyjay hit this problem when I was on a phone call with him. I wonder if he can comment if it’s still a problem. The workaround there was to make the AWS Portal Pulse the Primary. You can do that by right-clicking a Pulse entry in the Pulse panel while in super-user mode.
It looks like that issue I reported for that call is still open, so if the workaround ends up fixing it let me know and I’ll update the dev issue.
So the only times we have an issue with connecting to the slaves to send commends or get running logs its one of the following issues:
The launcher isn’t running on the slave.
The firewall is running and blocking communications with the server.
Our virus scanner is blocking communications with the server.
Pulse isn’t running (less likely).
Considering this is concerning cloud instances I’m not sure which of these would apply to that. We utilize an internal render farm. I hope some of that info helps.
