premature / false detection of test completion

[BenTheElder]

Running hydrohone@latest, testing Kubernetes 1.32:

$ hydrophone --conformance
21:53:15 INF API endpoint: https://34.31.43.236
21:53:15 INF Server version: version.Info{Major:"1", Minor:"32", GitVersion:"v1.32.0-gke.1358000", GitCommit:"526b8b157e735478210f48debfabd90b4c092acd", GitTreeState:"clean", BuildDate:"2024-12-11T22:26:48Z", GoVersion:"go1.23.3 X:boringcrypto", Compiler:"gc", Platform:"linux/amd64"}
21:53:15 INF Using namespace: conformance
21:53:15 INF Using conformance image: registry.k8s.io/conformance:v1.32.0
21:53:15 INF Using busybox image: registry.k8s.io/e2e-test-images/busybox:1.36.1-1
21:53:15 INF Test framework will start 1 thread(s) and use verbosity level 4.
21:53:15 INF Created Namespace conformance.
21:53:15 INF Created ServiceAccount conformance-serviceaccount.
21:53:15 INF Created Clusterrole conformance-serviceaccount:conformance.
21:53:15 INF Created ClusterRoleBinding conformance-serviceaccount-role:conformance.
21:53:16 INF Created Pod e2e-conformance-test.
21:53:16 INF Waiting up to 5m0s for Pod to start...
⠈⠁2024/12/17 21:53:21 Running command:
Command env: []
Run from directory: 
Executable path: /usr/local/bin/ginkgo
Args (comma-delimited): /usr/local/bin/ginkgo,--focus=\[Conformance\],--skip=,--no-color=true,--timeout=24h,/usr/local/bin/e2e.test,--,--disable-log-dump,--repo-root=/kubernetes,--provider=skeleton,--report-dir=/tmp/results,--kubeconfig=
2024/12/17 21:53:21 Now listening for interrupts
⠈⠑  I1217 21:53:22.908490      20 e2e.go:109] Starting e2e run "2157647e-6dba-49dc-a627-ecd863e96afb" on Ginkgo node 1
⢀⡱Running Suite: Kubernetes e2e suite - /usr/local/bin
====================================================
Random Seed: 1734472401 - will randomize all specs

Will run 411 of 6622 specs
⢎⡠22:39:31 INF Tests finished after 46m9s.
22:39:31 INF Downloading e2e.log to e2e.log...
22:39:31 INF Downloading junit_01.xml to junit_01.xml...
22:39:31 ERR Error: failed to download results: download failed: command terminated with exit code 1 (stderr: cat: can't open '/tmp/results/junit_01.xml': No such file or directory)

If I check the pod, the tests are definitely still running, which is why there's no junit yet:

$ kubectl get po -n=conformance e2e-conformance-test
NAME                   READY   STATUS    RESTARTS   AGE
e2e-conformance-test   2/2     Running   0          87m

$ kubectl exec -n=conformance e2e-conformance-test -c conformance-container -- cat /tmp/results/e2e.log | grep -o "•" | wc -l
373

# wait a bit ...

$ kubectl exec -n=conformance e2e-conformance-test -c conformance-container -- cat /tmp/results/e2e.log | grep -o "•" | wc -l
377

/kind bug
/sig testing

[BenTheElder]

I had a run on 1.32 without issues on another cluster (other than #223), so far I think the tests are passing and running normally, I can't tell why it decided to try to download the results early.