Is there any way to change the limits, open file descriptors in my case, both soft and hard, for a running process inside a pod?
I'm running a memcached deployment using helm stable/memcached, but the 1024 open file limit is really short for the intended concurrency.
If it is not possible to do so, what is the right way to change the limits for a deployment or globaly on a kubernetes cluster (running on AWS and setup with kops)?
The problem was that max simultaneous connections on memcached defaults to 1024, and the chart uses the defaults.
I needed to modify it to pass "-c "