Exporter process memory keeps increasing with time

[yashtessell]

I am not here to report an issue but I have more of a question about the oracle exporter's behavior. I am running the exporter in a pod/container and we are observing the exporter process memory keeps increasing with time (about 2MB every 5min).

I enabled pprof in the exporter and the profiler only indicates the exporter is taking 1-2MB of heap space. While top shows the process is taking ~45MB. Eventually, the pod reaches its allocated memory limits and restarts.

I noticed the memory suddenly jumps if a metric sweep from Prometheus comes in (request on /metrics endpoint). I am trying to make sense of it all and figure out why this is happening.

Any help is appreciated. Thanks.

For reference, I am including the output of pprof

Showing nodes accounting for 1447.25kB, 100% of 1447.25kB total flat flat% sum% cum cum% 902.59kB 62.37% 62.37% 1447.25kB 100% compress/flate.NewWriter 544.67kB 37.63% 100% 544.67kB 37.63% compress/flate.newDeflateFast (inline) 0 0% 100% 544.67kB 37.63% compress/flate.(*compressor).init 0 0% 100% 1447.25kB 100% compress/gzip.(*Writer).Write 0 0% 100% 1447.25kB 100% net/http.(*ServeMux).ServeHTTP 0 0% 100% 1447.25kB 100% net/http.(*conn).serve 0 0% 100% 1447.25kB 100% net/http.HandlerFunc.ServeHTTP 0 0% 100% 1447.25kB 100% net/http.serverHandler.ServeHTTP