Setup a render node in my corporate office, what is the best workflow or steps to take to access the render node with my client at my home office and utilize both machines?
Here's what I'd like to add. If you're connecting machines from two different physical locations, then setting up a VPN using something like hamachi or soft ether would be a good idea.
Crowdrender works best on a local network, but you can also use it on a virtual private network, which, when properly setup, looks pretty much the same as far as the software can tell. If you have a corporate VPN client, then you can probably just use that.
Be aware that with some VPN software, ALL your computers have to join the same VPN to be able to connect to each other as some VPN solutions can isolate your computer from your local network once attached to the VPN.
Another consideration is to have realistic expectations of the transfer speeds and latency when connecting peer to peer. We've done tests between locations that are not too far away geographically, but the tests showed that there was a LOT of latency and as a result times to upload projects were larger than using a render farm service where the project had to cross the pacific!
This is not surprising if you look at how many hops there are between your network and the network you are trying to reach. For example, if you run trace route, like we did, the number of hops between my and chris's house was 23, quite a lot given on a clear day I can see his house.
I did the same to google and it came out at 8 hops. A hop just means there is a computer between my and Chris's local networks. More computers = more delay and slower transfers generally speaking.
Hopefully this gives you some idea of what to expect.
Just make sure you set up the docker container for your needs, I personally added parameters to autostart the container on boot, my secondary machine is just running ubuntu and followed the same instructions for the Nvidia cards support
Just to let everyone know, there is a PR to make some serious changes to the docker containers. Anyone interested can read it here - https://github.com/crowdrender/cr-docker/pull/21
between the 3 office locations, we are all using the unifi ecosystem. ill look into the vpn setup, thanks guys
First, thanks to @charmandrigo for that suggestion.
Here's what I'd like to add. If you're connecting machines from two different physical locations, then setting up a VPN using something like hamachi or soft ether would be a good idea.
Crowdrender works best on a local network, but you can also use it on a virtual private network, which, when properly setup, looks pretty much the same as far as the software can tell. If you have a corporate VPN client, then you can probably just use that.
Be aware that with some VPN software, ALL your computers have to join the same VPN to be able to connect to each other as some VPN solutions can isolate your computer from your local network once attached to the VPN.
Another consideration is to have realistic expectations of the transfer speeds and latency when connecting peer to peer. We've done tests between locations that are not too far away geographically, but the tests showed that there was a LOT of latency and as a result times to upload projects were larger than using a render farm service where the project had to cross the pacific!
This is not surprising if you look at how many hops there are between your network and the network you are trying to reach. For example, if you run trace route, like we did, the number of hops between my and chris's house was 23, quite a lot given on a clear day I can see his house.
I did the same to google and it came out at 8 hops. A hop just means there is a computer between my and Chris's local networks. More computers = more delay and slower transfers generally speaking.
Hopefully this gives you some idea of what to expect.
I personally followed this:
Just make sure you set up the docker container for your needs, I personally added parameters to autostart the container on boot, my secondary machine is just running ubuntu and followed the same instructions for the Nvidia cards support