Hi All, Our current embedded Docker container works best when run from Docker with the "-it" (interactive) flag. Since K8s pods are seldom interactive, it seems we'll need to create a container focused on the needs of server deployment. DRILL-7563 [1] has an outline of the tasks. Fortunately, the steps are pretty standard; many projects have led the way.
A first-draft Drillbit container is hosted at [2], source code is at [3]. (These are personal repos for now.) The basics work under Docker: I can connect to the Drill Web Console and run a query. The key dependency is ZK [4]. The next step is to create a sqlline container that can be run in interactive mode and which connects to the containerized Drillbit(s). K8s has nice ways to pass configuration into a container, which we'll need to run the Drillbit with other than the out-of-the-box configuration. (Memory will be the first thing someone wants to change.) Rather than fight with the clunky Docker solutions to config, I'll just jump directly over to the K8s approach. Suggestions, feedback and help welcome. Thanks, - Paul [1] https://issues.apache.org/jira/browse/DRILL-7563 [2] https://hub.docker.com/repository/docker/gaucho84/drill [3] https://github.com/paul-rogers/drill-docker See the docker folder for details. [4] https://hub.docker.com/_/zookeeper/ On Friday, January 31, 2020, 9:21:49 PM PST, Ron Cecchini <[email protected]> wrote: FWIW, I found a thread on SO that addresses this OpenShift issue with containers seemingly exiting for no apparent reason, even when running as root. It might be TTY related. openshift pod fails and restarts frequently https://stackoverflow.com/questions/35744404/openshift-pod-fails-and-restarts-frequently/36234707#36234707 What is CrashLoopBackOff status for openshift pods? https://stackoverflow.com/questions/35710965/what-is-crashloopbackoff-status-for-openshift-pods > On January 31, 2020 at 4:40 PM Ron Cecchini <[email protected]> wrote: > > > Thank you, Paul, for your in depth and informative response. > > Here's what we did today as a test: > > In OpenShift, we enabled allowing containers to run however they wanted and > redeployed the Drill Docker. [...] > The result is that the container comes up - shows that the user.name is > indeed root - doesn't give any errors ... > > And then seemingly goes away, again w/o any indication of an error. And then > OpenShift goes into its crash-reboot cycle trying to restart the container, > and each time we see the same thing:
