Parsing large OpenAPI documents leads to memory issues

[jcmcken]

E.g. if you parse the Kubernetes swagger.json (~5.1MB of JSON as of this issue), even with a pretty beefy machine, you eventually run out of memory.

[andreb89]

I had the same issue. The solution is to install pip with python3 and NOT python2. The memory leak issue only happens, at least for me, if I install openapi2jsonschema with the python2 pip.