kola-denylist: use string instead of list for tracker

[miabbott]

I strongly believe that I tested the use of a list in the denylist as
part of #595, but evidence says otherwise.

[miabbott]

[coreos-assembler]$ (cd src/config/ && git show HEAD)                                                                                                           
commit c4cef1eea5112b7197b554f4f9a70f29beb4224c (HEAD -> fix_denylist, miabbott/fix_denylist)
Author: Micah Abbott <miabbott@redhat.com>
Date:   Wed Jul 28 15:34:27 2021 -0400
                                                                               
    kola-denylist: use string instead of list for tracker
                       
    I strongly believe that I tested the use of a list in the denylist as
    part of #595, but evidence says otherwise.
                                                                               
diff --git a/kola-denylist.yaml b/kola-denylist.yaml
index 7beb6ac..34875e4 100644                   
--- a/kola-denylist.yaml               
+++ b/kola-denylist.yaml          
@@ -6,9 +6,7 @@       
   arches:                  
    - s390x            
 - pattern: crio.base           
-  tracker:                                                         
-    - https://github.com/kubernetes/kubernetes/issues/87325a
-    - https://bugzilla.redhat.com/show_bug.cgi?id=1986239
+  tracker: "https://github.com/kubernetes/kubernetes/issues/87325 && https://bugzilla.redhat.com/show_bug.cgi?id=1986239"
 # for s390x by-partlabel can't be used and even if that is avoided by using part-uuid, still depends on the cpi fix below
 - pattern: ext.config.var-mount
   tracker: https://github.com/ibm-s390-tools/s390-tools/pull/82
[coreos-assembler]$ jq . < /cosa/coreos-assembler-git.json 
{
  "date": "2021-07-28T14:30:42Z",
  "git": {
    "commit": "8d633557ccda112ceea30bc3462180256af09013",
    "origin": "git@github.com:coreos/coreos-assembler.git",
    "branch": "HEAD",
    "dirty": "false"
  },
  "file": {
    "checksum": "53bc7e3b05fb4d4698a9be84939bfe67d22468b66e6c86cdb1584f05e4ab55d4",
    "checksum_type": "sha256",
    "format": "tar.gz",
    "name": "coreos-assembler-git.tar.gz",
    "size": "39437099"
  }
}
[coreos-assembler]$ kola run 
⚠️  Skipping kola test pattern "crio.base":
  👉 https://github.com/kubernetes/kubernetes/issues/87325 && https://bugzilla.redhat.com/show_bug.cgi?id=1986239
⚠️  Skipping kola test pattern "coreos.ignition.journald-log":
  👉 https://github.com/coreos/coreos-assembler/issues/1173
=== RUN   coreos.tls.fetch-urls
=== RUN   ext.config.root-reprovision.raid1
=== RUN   coreos.boot-mirror.luks
=== RUN   rpmostree.status
=== RUN   coreos.ignition.failure
=== RUN   luks.sss.t1
=== RUN   rhcos.network.bond-with-dhcp
=== RUN   coreos.ignition.once

[cgwalters]

I strongly believe that I tested the use of a list in the denylist as part of #595, but evidence says otherwise.

You mean you can't confirm...or deny it.

[cgwalters]

If this fails let's merge anyways but I want to see where we get from this run - at least kata should be in now.

[miabbott]

error: Packages not found: bootupd, bootupd

This is a comedy of errors

[cgwalters]

aos repo just got it, not sure if there's another process needed to sync
/retest

[cgwalters]

/override ci/prow/build-test-qemu
/lgtm

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: cgwalters, miabbott

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [cgwalters,miabbott]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[openshift-ci]

@cgwalters: Overrode contexts on behalf of cgwalters: ci/prow/build-test-qemu

In response to this:

/override ci/prow/build-test-qemu
/lgtm

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[miabbott]

aos repo just got it, not sure if there's another process needed to sync
/retest

Perhaps the sync from internal to the CI cluster is on a timer or what not

[jlebon]

If you somehow used cosa before coreos/coreos-assembler#2296 when testing #595, that could explain why it worked. In Python, we just stringified it when printing. But we lost that when moving to stronger typing in golang.

Not a bad idea to officially support it though, even if just as a different e.g. trackers key.