Separate container/vm for shell access

Hi there,

I’ve recently launched a Rails app in MVP mode, and since it’s still small I’m able to get by with just the starter plan and a single node. Normally this app uses around ~350-400mb of RAM which is expected. So far the experience with Render has been great!

It seems that anytime I go into shell mode and try to start a rails console to look at some data, I can see the memory usage double (expected), but then it seems that the node gets terminated or something. My website becomes unresponsive, my shell access gets terminated, and I then see the application reboot. Basically means I can’t really use the shell for much.

One great thing about the heroku console is that it launches a separate container/dyno, which is then billed by the second. Super helpful for rails apps so I can easily explore the data, or do small data tweaks and things like that.

Is something like this possible with Render?

Hello Mike,

One thing that you could try is create an additional Web service from the same repository and use that for shell access. You can then suspend this additional service when not in use.