ariadne/tests/test_hermes_storage_signals.py
codex ee7b243106
All checks were successful
Tests / Declarative: Post Actions passed: 1257
feat(hermes): clear stuck agent pods, and teach Hermes the new remediations
A build whose agent never started is a distinct failure from one that lost a
connection mid-run: retrying can work, but when the pool is already full of
stuck pods the retry queues behind them and fails identically. Clearing first
is what makes the retry worth making. The clear is Ariadne's existing
scheduled pod cleanup, which only removes pods that have already succeeded or
failed, so nothing running is touched. This is the failure behind lesavka's
open issue and behind two stalled demo runs tonight.

Critically, all three remediations are now described in the triage prompt.
They were wired in Ariadne but absent from what Hermes is told, so Hermes
could never have requested them - the allowlist would have advertised
capability that could not fire. A test now asserts every allowlisted action id
appears in the prompt, so the two cannot drift apart again.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-06 21:04:38 -03:00

75 lines
2.6 KiB
Python

"""Tests for recognising Jenkins workspace storage exhaustion."""
from __future__ import annotations
from ariadne.services import hermes_storage_signals as module
def _bundle(*lines: str, tail: str = "") -> dict:
return {"jenkins": {"console_failures": [{"text": "\n".join(lines)}], "console_tail": tail}}
def test_a_full_workspace_volume_is_recognised() -> None:
matched, marker = module.has_storage_exhaustion_signature(
_bundle("cp: error writing '/home/jenkins/agent/workspace/x': No space left on device")
)
assert matched is True
assert marker == "no space left on device"
def test_the_tail_is_searched_too() -> None:
matched, _ = module.has_storage_exhaustion_signature(
_bundle("nothing here", tail="workspace volume full: disk quota exceeded")
)
assert matched is True
def test_a_full_disk_without_a_workspace_hint_is_not_this_failure() -> None:
"""Reclaiming Jenkins workspaces would not address it."""
matched, marker = module.has_storage_exhaustion_signature(
_bundle("etcd: No space left on device")
)
assert (matched, marker) == (False, None)
def test_a_workspace_mention_without_a_storage_marker_does_not_match() -> None:
matched, _ = module.has_storage_exhaustion_signature(_bundle("workspace checkout failed"))
assert matched is False
def test_never_raises_on_hostile_input() -> None:
for bad in (None, 7, {}, {"jenkins": None}, {"jenkins": {"console_failures": "no"}}):
assert module.has_storage_exhaustion_signature(bad) == (False, None)
def test_agent_provisioning_failures_are_recognised() -> None:
"""The build never got an agent, rather than failing once running."""
from ariadne.services import hermes_infra_signals
for text in (
"All nodes of label 'lesavka_583-abc' are offline",
"Error in provisioning; agent=KubernetesSlave[x]",
"Still waiting to schedule task",
):
matched, marker = hermes_infra_signals.has_agent_provisioning_signature(_bundle(text))
assert matched is True, text
assert marker
def test_an_ordinary_failure_is_not_an_agent_problem() -> None:
from ariadne.services import hermes_infra_signals
matched, _ = hermes_infra_signals.has_agent_provisioning_signature(
_bundle("AssertionError: expected 1 got 2")
)
assert matched is False
def test_agent_signature_never_raises() -> None:
from ariadne.services import hermes_infra_signals
for bad in (None, 7, {}, {"jenkins": {"console_failures": "no"}}):
assert hermes_infra_signals.has_agent_provisioning_signature(bad) == (False, None)