So it turns out that memory wasn't actually growing. Instead, the openshift
deployment we were using was based on a very low metaspace sizing with a hard
upper limit. This was missed, and I'm not sure why they put this limit in but
they do. We see steady memory usage now after giving it a
Hi,
I find strange that memory usage grows. Could you provide a short code snippet
how you invoke the client?
As a solution: in most of cases the client proxies are thread safe and can be
reused for concurrent calls - it can improve performance and memory in your
case.
If you are applying
