ci(e2e): raise inotify limits for ubuntu-26.04 runners

The ubuntu-latest runner image rolled over to ubuntu-26.04, where
/etc/sysctl.conf is no longer applied at boot (actions/runner-images#15108).
This drops fs.inotify.max_user_instances from the image's intended 1280 to
the kernel default of 128. inotify limits are not namespaced, so the 3-node
kind cluster used by the e2e tests exhausts them and Tekton entrypoint pods
fail with "failed to create fsnotify watcher: too many open files", making
the pipelinerun/taskrun log assertions fail.

Restore the intended limits on the runner host before creating the kind
cluster.

Assisted-by: Claude Opus 4.8 (via GitHub Copilot CLI)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Signed-off-by: divyansh42 <diagrawa@redhat.com>

Author: 2 people

.github/workflows/e2e-matrix.yml

@@ -51,6 +51,18 @@ jobs:
 
         echo "${GOPATH}/bin" >> "$GITHUB_PATH"
 
+    - name: Increase inotify limits
+      run: |
+        # Workaround for actions/runner-images#15108: on the ubuntu-26.04
+        # runner image /etc/sysctl.conf is no longer applied at boot, so
+        # fs.inotify.max_user_instances falls back to the kernel default of
+        # 128 (instead of the image's intended 1280). inotify limits are not
+        # namespaced, so the 3-node kind cluster exhausts them and Tekton
+        # entrypoint pods fail with
+        # "failed to create fsnotify watcher: too many open files".
+        sudo sysctl -w fs.inotify.max_user_instances=1280
+        sudo sysctl -w fs.inotify.max_user_watches=655360
+
     - name: Run tests
       run: |
         ./hack/setup-kind.sh \