Skip to content

Commit c1fe1e9

Browse files
committed
ci(e2e): say what the 215-minute e2e budget actually covers
The comment on every lane claimed the budget fits two tenant Kubernetes suites failing at their node-join deadline. A green full suite on the container lane already spends most of it, and a suite that runs into the deadline costs about 25 minutes more than one that passes, so the claim does not hold. State the real bound so a cancelled run is read as a red suite, not as a budget to raise. Assisted-by: LLM Signed-off-by: Aleksei Sviridkin <[email protected]>
1 parent da28ea7 commit c1fe1e9

4 files changed

Lines changed: 12 additions & 8 deletions

File tree

‎.github/workflows/e2e-fork.yaml‎

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -540,8 +540,9 @@ jobs:
540540
# per-node 8 CPU / 24 GiB in hack/e2e-compose.yaml is why the class is this
541541
# one. See the `e2e` job there for the full sizing note.
542542
runs-on: oracle-vm-32cpu-128gb-x86-64
543-
# Sized above what the tenant-Kubernetes pair can spend on a failing run, and
544-
# held to one figure across every lane that runs the suite.
543+
# Covers a green full suite. It does not cover a tenant-Kubernetes suite
544+
# running into its node-join deadline, so a red there can end as a
545+
# cancelled job. Held to one figure across every lane that runs the suite.
545546
timeout-minutes: 215
546547
permissions:
547548
contents: read

‎.github/workflows/e2e-tag.yaml‎

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -134,8 +134,9 @@ jobs:
134134
# leaves enough host headroom for networking, containerd, and the runner
135135
# agent while the full platform installs across all three nodes.
136136
runs-on: oracle-vm-32cpu-128gb-x86-64
137-
# Sized above what the tenant-Kubernetes pair can spend on a failing run, and
138-
# held to one figure across every lane that runs the suite.
137+
# Covers a green full suite. It does not cover a tenant-Kubernetes suite
138+
# running into its node-join deadline, so a red there can end as a
139+
# cancelled job. Held to one figure across every lane that runs the suite.
139140
timeout-minutes: 215
140141
permissions:
141142
contents: read

‎.github/workflows/nightly.yaml‎

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -247,8 +247,9 @@ jobs:
247247
# containerd and the runner agent. Enforced by
248248
# hack/sandbox-runner-headroom.bats.
249249
runs-on: oracle-vm-32cpu-128gb-x86-64
250-
# Sized above what the tenant-Kubernetes pair can spend on a failing run, and
251-
# held to one figure across every lane that runs the suite.
250+
# Covers a green full suite. It does not cover a tenant-Kubernetes suite
251+
# running into its node-join deadline, so a red there can end as a
252+
# cancelled job. Held to one figure across every lane that runs the suite.
252253
timeout-minutes: 215
253254
permissions:
254255
contents: read

‎.github/workflows/pull-requests.yaml‎

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -939,8 +939,9 @@ jobs:
939939
# scheduler and not what a pod reads from /proc, so a workload that sizes
940940
# itself from /proc/cpuinfo or /proc/meminfo is configured for the whole host.
941941
runs-on: oracle-vm-32cpu-128gb-x86-64
942-
# Sized above what the tenant-Kubernetes pair can spend on a failing run, and
943-
# held to one figure across every lane that runs the suite.
942+
# Covers a green full suite. It does not cover a tenant-Kubernetes suite
943+
# running into its node-join deadline, so a red there can end as a
944+
# cancelled job. Held to one figure across every lane that runs the suite.
944945
timeout-minutes: 215
945946
permissions:
946947
contents: read

0 commit comments

Comments
 (0)