We have been seeing a good number of CI test cases failing. This does become a problem with taking in new fixes and having them marked as bad tests reduces the test coverage. As a result, we are planning to have a hackathon on 09th April. This will be a virtual one happening on google meet. Information to join can be found below.
We will be working on the tests that are currently failing spuriously and also on the test cases that have been marked as bad in the past and fix these. The consolidated list  has all the test cases we will look into during this hackathon. If any of you have come across a test which has failed and was missed in the list, feel free to link the test case and the link to failure in the consolidated list .
Prerequisite would be to have a centos machine where you can clone gluster and work on.
The action item would be to take up as many test cases as possible(write down your name against the test case to avoid rework), run it on the local environment, find out why it fails, send out a fix for it and review others' patches. If you want to add more test cases as per your usage, feel free to do so.
You can go through the link to get the basic idea on how to set up and contribute. If there are more questions, please feel free to ask them, also you can discuss it with us during the hackathon and sort it as well.
We will send out a google calendar invite soon. The tentative timing will be from 11am to 4:30pm IST
Recently I've reported https://github.com/gluster/glusterfs/issues/1148 and now trying to do some debugging. I'm starting with client (e.g. qemu-img)
backtraces which exposes some EC innards but seems doesn't show anything
related to sharding. So, any ideas and suggestions around further debugging
would be very helpful.