cc: @cverna
https://koji.fedoraproject.org/koji/taskinfo?taskID=33189799
image build failed: {"aarch64": {"general": "OsbsException('ConnectionError(ReadTimeoutError(\"HTTPSConnectionPool(host=\\'osbs-aarch64-master01.arm.fedoraproject.org\\', port=8443): Read timed out.\"))')"}}
Metadata Update from @cverna: - Issue assigned to cverna
For the time being I disable the cluster in
https://infrastructure.fedoraproject.org/cgit/ansible.git/commit/?id=de92d34f55135e9c9d448e10616354d64a5935ef
I ll keep the ticket open to investigate why this is still failing.
Metadata Update from @mizdebsk: - Issue priority set to: Waiting on Assignee (was: Needs Review)
Any news here @cverna ?
So yes the aarch64 cluster does not work well. The docker-containerd daemon crashes after 1 hour or so. I have not have much time to look at it in more details but I suspect the fact we have only one box that runs both the OpenShift master and node does not help.
@kevin do you know if we will have more aarch64 boxes available ? or maybe we should look at using some AWS resources ?
The problem with AWS resources is that all the things you want to work on have to be copied out of PHX2, stored in AWS, worked on in the AWS servers and then copied back down to PHX2. That isn't fast or free. The ways to make it better quickly turn into a huge 6 to 18 month project as you are rewriting all of infrastructure. We have been asked not to do things like that without plans.. so we need to do that first.
Good news though, we have aarch64 boxes for you.
Going to close that, since we are redeploying OSBS and will more likely have more aarch64 boxes.
Metadata Update from @cverna: - Issue close_status updated to: Fixed - Issue status updated to: Closed (was: Open)
Login to comment on this ticket.