Memory Management on Low Spec Systems

[dsriseah]

I created a basic Digital Ocean droplet, which is 512MB with 1 shared vCPU, and installed URSYS on it. The install went pretty cleanly, but running the ur net start command failed. It was extremely slow, and I am suspecting that it's running out of memory or some other resource.

I need to add ways of monitoring memory usage and ensure that the way I'm forking commands is not as heavy-weight as it currently is. I probably need to reduce the number of node scripts that are launching other node scripts, and rely more on the bash scripting side of things.

See URSYS Wiki for debugging process notes.

Things to Try

• Add memory heap instrumentation to the node scripts themselves
• Use the free -h command to see what available memory is and how it shrinks.

[dsriseah]

I was mistaken about the minimum DO droplet size: it's been 1GB all this time. Otherwise, the server just runs out of memory. As an additional note, to use the Visual Studio Code remote server requires an additional 1GB of memory! Otherwise, the server just bogs down and crashes.

I've updated my test instance to 2GB ($12/month from $6/month) to see if it makes a difference.