Last week I attempted to debug the server to find the memory leak with the git application. TLDR: Could not find it. It was literally a pile a spaghetti. So retaining the knowledge from trying to solve problems with their Github, I just remade an entire node from scratch combining multiple tutorials. These normally were with separate elements [ie. node + mongo, node + socket, node + request, request + python] just merged together with a little logic.
So the overall results were great, not pretty, but hey at least it doesn’t crash anymore!
Oh wow look! so much colour and excite! It is currently working with just over
800,000 with the inefficient query. It searching finishes and doesn’t have a constant memory gain overtime so it doesn’t crash. There are some problems with this server as it all can’t be uphill:
- When uploading lost of data via requests, it seems to cap at 25% CPU, stops and waits about 5s, then continues.
- The query takes forever to load because it caps at 100% CPU, but does it in stages
- This also stops the uploading
There is probably a cap on the way node is done. Also I am currently using the smallest AWS server which is super slow and apparently has the processing power of a 2006 chip. Some more investigating is needed to fix this, I am going to ask Jorke about the potential problems and try to craft some solutions.