Raki

Hey @pydctw @joyvuu-dave - do you think it's realistic to expect a maintainer to pick this up anytime soon? If not, would you be willing to accept PRs? The issue...

@mtulio - that would be amazing mate, it would avoid dupe work on my team's side. We heavily use Sonobuoy here at Microsoft for our Kubernetes CI testing, I'd be...

@mtulio - just a heads up, I tried basically what's in your PR and in the Bash above. One of our tests generate about 8 GB of logs, and unfortunately...

@mtulio Ah that's clever, you're directly finding the node that runs the aggregator and pulling it out. The problem is....we run Sonobuoy every night as part of our CI pipelines...

Here's a solution, just hack out the lines in the [rest.py](https://github.com/kubernetes-client/python/blob/5a96bbcbe21a552cc1f9cda13e0522fafb0dbac8/kubernetes/client/rest.py#L232) file that is the root of all evil in a dockerfile during build: ```bash find /usr/local/lib/python*/dist-packages/kubernetes/client/rest.py -type f -exec...

Hey @roycaihw - we ran a benchmark on our pod to quantify the memory leak from this bug if you run an infinite loop for 5 minutes calling K8s API...

Tagging @yliaog @kariya-mitsuru @palnabarun as original authors of [rest.py](https://github.com/kubernetes-client/python/blob/5a96bbcbe21a552cc1f9cda13e0522fafb0dbac8/kubernetes/client/rest.py)

> @mdrakiburrahman what's the memory usage if you apply that patch vs without it? With the hack patch, about ~200 MB - steady, as expected for my pod. Without the...

With hack:  Without hack: 

@ebattat the issue is, `sqlservr` is not able to boot up due to `restricted-v2` dropping `ALL`, which also drops `NET_BIND_SERVICE`, which `sqlservr` uses to bind the default MSDTC RPC on...