NFD

Note: if you plan to use Virtualbox for these agents, you'll have to upgrade to version 6.1.14 or later in order to have support for the new Ubuntu kernel. And that in turn implies upgrading Vagrant to 2.2.7 or later (the latest is 2.2.10 and has been working fine for me so far).

Noted. But I think Saurab will need to take care of this one. UA still does not work with the git issue originating in 20.04.

Md Ashiqur Rahman wrote in #note-2:

UA still does not work with the git issue originating in 20.04.

Maybe that will no longer be an issue in 20.10... I hope.

What's the status here?

Waiting for bento box. The generic one came out just two days ago, I can give it a try tomorrow.

Davide Pesavento wrote in #note-1:

Note: if you plan to use Virtualbox for these agents, you'll have to upgrade to version 6.1.14 or later in order to have support for the new Ubuntu kernel. And that in turn implies upgrading Vagrant to 2.2.7 or later (the latest is 2.2.10 and has been working fine for me so far).

Ok, I overlooked this comment and set up the VM with VirtualBox ver 6.1.2 and vagrant ver 2.1.6. I will need to upgrade these two first tomorrow before the deployment then.

I added two 20.10 nodes on maestro but when I run a test with refs/heads/master for NFD, I'm getting the following git timeout error:
https://jenkins.named-data.net/job/NFD/7884/console

I see a NullPointerException there, not a timeout. But it seems to be happening on other nodes and other projects as well. And I remember seeing this in the past too. It sounds like a jenkins bug to me.

Ah, yes. You're right. But then I'm confused on why this is happening when I trigger with the master branch but works fine when someone pushes a commit/build. I'm also forwarding this to Eric to help out.

No idea. In any case, I think we should report the NullPointerException upstream. Even if there's a misconfiguration somewhere, it shouldn't crash.

I reported the error upstream and a fix has been merged. It will appear in the next release of the git plugin.

In the meantime I found a workaround by setting the "Branches to build" to $GERRIT_BRANCH instead of ** (any).