server restarts when K8s pod workspaces are started

[ericpaulsen]

both @sharkymark and I have observed occasional restarts of the coder-logstream-kube deployment when it tries to stream startup events of workspace pods. this does not happen when a workspace is built as a deployment.

i've tailed the logs during a pod build, and grep-ed "error":

fatal error: stack overflow

[sharkymark]

Confirming I experience this.

This should be prioritized since restarting pods could impact other workspaces depending on this logstream feature.

[kylecarbs]

Oofta. Fixing now.