Issue: 

Getting profiling data in the file system of the pod doesn't seem very flexible for a customer. We need persistent volume for the profiling data and an efficient way to access the profiling data for a customer.

Questions

1. What approaches or strategies have we already discussed earlier and didn't work out for us?
2. The profiling data in the container is getting overrode with the new profiling data or a new file is created for the new profiling data? if yes, then what about data cleanup? are we setting some threshold for storing the new profiling data files and removing the old ones?
3. Are we completely losing the profiling data files once the pod is recreating or getting deleted?

Some Fuzzy Approaches for PV:

1. If we create some master pod that will poll the profiling data from all the other pods? What about data redundancy? what will be the cleanup strategy if we gonna store all the profiling data files from all the pods? what if we only override the latest profiling data from each pod into the master pod's file system? would it be enough for the customer? How the networking will be integrated into this scenario?
2. Another better approach?