Replace the non-existent myrepo.git placeholder (which fails with
"repository not found") with this repo's URL and a harmless
`echo build-ok` build command, so §3 and §4.3 run as pasted.
The `job` command enforces --template-path (job.py:447) and defaults
--snapshot-name to BaseClean, which is wrong for Linux. The §3 example
omitted both and failed with "--template-path is required". Add both
flags (BaseClean-Linux) and note them as required.
Replace the terse, abbreviation-heavy style with clean prose and
consistent formatting. Make command blocks copy-paste-ready on this host:
capture the guest IP into a shell variable for `validate guest` / `build
run`, expand globbed VMX paths, and group the `job` flags into readable
lists. Add intra-doc anchor links throughout. Point the pwsh-rollback note
at scripts/archived/. No commands or flags changed.
The shim mapping table now notes the .ps1 live in scripts/archived/
(frozen, lint-exempt, no live caller) and gains the Phase C host tools
(bench/smoke/validate host/creds set). Fix the "stays in PowerShell"
list: drop Test-CapacityBurnIn (archived) and the non-existent
Measure-CIBenchmark; keep retention/backup/template/e2e/Windows-side.
Document `vmrun getGuestIPAddress` (run as ci-runner) for templates and
clones as the prerequisite for `validate guest`, with the AGENTS.md #1
"Error: on stdout" gotcha. Update the validate-guest Linux example to the
live template IP.
Move 16 .ps1 files to scripts/archived/ — they have no live caller:
- 10 pure shims (New-BuildVM, Remove-BuildVM, Cleanup-OrphanedBuildVMs,
Wait-VMReady, Invoke-RemoteBuild, Get-BuildArtifacts, Watch-DiskSpace,
Watch-RunnerHealth, Get-CIJobSummary, Invoke-CIJob) that only forwarded
to `ci_orchestrator`; the live Gitea composite action now calls
`python -m ci_orchestrator job` directly.
- 6 Phase-C host tools reimplemented natively in Python (Start-BurnInTest
[+ -Linux], Test-CapacityBurnIn → bench; Test-Smoke → smoke run;
Validate-HostState → validate host; Set-CIGuestCredential → creds set).
Add scripts/archived/README.md mapping each file to its `ci` replacement,
and exempt scripts/archived/ from the PSScriptAnalyzer lint job.
Kept (still used / not yet ported): Register-CIScheduledTasks,
Backup-CITemplate, Invoke-RetentionPolicy, Set-TemplateSharedFolders,
Test-CIGuestWinRM, the e2e build harnesses, and all template/ guest
provisioning scripts.
§4.7 now explains the grant is per-session and points to setup Step 8 /
the standalone install-vmware-gui-access.sh, keeping the manual one-off
xhost command as a fallback.
`ci vmware` runs as ci-runner and relies on host-based X auth (xhost),
but the grant is per-session and wiped on logout / X restart, so the GUI
launch fails with "cannot open X display" after every desktop relogin.
Add scripts/install-vmware-gui-access.sh: installs an XDG autostart entry
in the desktop user's home that re-applies `xhost +SI:localuser:ci-runner`
at each graphical login, and applies it immediately when a display is
reachable. Desktop user resolves from --desktop-user / $CI_DESKTOP_USER /
$SUDO_USER (never root); service user from $CI_RUN_USER. --uninstall
removes the autostart entry and revokes the live grant.
Wire it into setup-host-linux.sh as Step 8, with --desktop-user (default
$SUDO_USER) and --skip-gui-access for headless hosts.
The GUI connects via the X host ACL (xhost), not the inherited XAUTHORITY
cookie (unreadable by ci-runner). On a normal session the grant is already
present and `ci vmgui` works with no --display/xhost dance — sudo forwards
DISPLAY. xhost is only needed if you hit "cannot open X display". Drop the
overstated "--display :0 required" framing from §4.7.
`vmgui` sent the child's stderr to /dev/null, so when the X display was
unreachable (the common case: sudo -u ci-runner loses the desktop user's
XAUTHORITY and xhost wasn't granted) the GUI died with "cannot open display"
and the command still printed success — "nothing opens".
Now it:
- pre-flights the display with xdpyinfo and aborts with the exact
`xhost +SI:localuser:ci-runner` hint instead of spawning a doomed process;
- drops an inherited XAUTHORITY that this user can't read (falls back to
host-based xhost auth);
- logs the GUI's stdio to /tmp/vmgui-*.log and grace-polls 1.5s, surfacing the
log if the GUI exits immediately;
- errors (non-zero) when no DISPLAY is resolvable.
Runbook §4.7 updated with the "if nothing opens" guidance. 13 tests; green ≥90%.
Add top-level `vmgui` to open the VMware Workstation GUI (renamed from the
earlier `vm open`). --vmx is optional: with it the GUI opens that VM, without
it the GUI starts bare. Power-on/fullscreen flags ignored (with a notice) when
no VMX is given. Needs an X display — sudo -u ci-runner strips $DISPLAY, so
--display / xhost; see runbook §4.7. 9 tests; suite green ≥90%.
Rewrite docs/RUNBOOK-PhaseC.md as a full manual-command reference: every
ci_orchestrator command with a working example as the ci-runner user, the `ci`
alias setup, quick-reference table, manual VM lifecycle (incl. vm open GUI
launch with the xhost/--display procedure), templating, maintenance, Phase C
testing, cutover checklist and a troubleshooting map.
scripts/install-ci-alias.sh installs `ci` (runs the orchestrator as ci-runner
via the prod venv) and `ci-me` (current user). Idempotent, marker-bounded;
supports --system (/etc/profile.d), --uninstall. setup-host-linux.sh gains a
step [7/7] that installs it system-wide, passing CI_PROD_PYTHON/CI_RUN_USER.
Open a VMX in the interactive Workstation GUI (not the headless vmrun path) for
hands-on debugging of a clone. Spawns `vmware` detached (own session, stdio
detached) and returns the child PID.
Flags: --vmx (required), --power-on (-x), --fullscreen (-X), --new-window (-n),
--display (sudo -u ci-runner strips $DISPLAY, so allow forcing it),
--vmware-path. Warns when no X display is set. 7 tests; vm.py 96%.
- CLAUDE.md: Linux host no longer runs pwsh; PS 5.1 constraints scoped to
Windows guests + Windows runner.
- RUNBOOK: add "Phase C — Python equivalents" table; §6-§13 benchmark records
kept as historical provenance.
- TODO §3.7: close $IsWindows shim bug (resolved-by-replacement — bench run
bypasses the shim).
- plans: mark C1-C8 done + docs C9; flag operational cutover/uninstall pending.
Apply analysis §4 option 1: pssa job runs on windows-build with
shell: powershell, gated to workflow_dispatch so push/PR don't queue it when
no Windows runner is registered. The python job stays on linux-build. No job
on a Linux runner uses pwsh anymore.
Remove _Common.psm1, _Transport.psm1 and their Pester test, plus their sole
consumer Measure-CIBenchmark.ps1 — all replaced by `bench measure`. Leaving
the script would dangle an Import-Module on a deleted .psm1.
Add four native command groups so the Linux host no longer needs pwsh for
manual ops. __main__ registers all four at once, so they ship together.
- bench run (C2): N concurrent jobs in-process via concurrent.futures,
per-round orphan-clone + stale-lock assertions. Replaces
Test-CapacityBurnIn.ps1 + Start-BurnInTest*.ps1. No
Start-Job/pwsh/$IsWindows.
- bench measure (C3): clone/start/IP/transport/destroy timings appended to
benchmark.jsonl (legacy field names kept). Replaces
Measure-CIBenchmark.ps1.
- smoke run (C4): one E2E job; asserts exit 0 + artifact dir + job/success
event. ns7zip/nsinnounp presets. Replaces Test-Smoke.ps1
+ Test-*Build-Linux.ps1.
- validate host (C5): vmrun/snapshot/keyring/permissions/systemd checks.
- validate guest(C6): transport probe with explicit failure cause.
- creds set (C6): keyring writer matching credentials.py read scheme;
password via stdin/prompt only. Replaces
Set-CIGuestCredential.ps1.
All groups import backends.load_backend only (Phase D ESXi path stays open).
Suite green, coverage 95.10% (gate 90%); new modules at 100%.
Phase B is closed (checklist Passi 1–9 [x], benchmark matrix §6–§14
recorded). Move the two finished working docs into the dated archive:
plans/PhaseB-user-checklist.md → plans/archived/2026-06-07/
plans/benchmark-windows-host.md → plans/archived/2026-06-07/
plans/ now holds only the forward-looking plans (idea-3 Phase C pwsh
removal, idea-4 Phase D ESXi, ideas-overview). Fixed the live links that
pointed at the moved/relocated files: idea-3 §4, idea-2 §8 (relative path),
and ideas-overview (idea-1/idea-2 now under archived/, Phase A+B marked ✅).
RUNBOOK §15: dual-boot operation — Linux⇄Windows switch procedure, per-OS
CI-stack table (roots/runner/vmrun), current GRUB behaviour (DEFAULT=0,
TIMEOUT=0 hidden → boots straight to Linux), post-boot verification commands
for both OSes, and an optional GRUB tweak (left unapplied — boot-to-Linux is
intentional so the box returns as the Linux CI host after any unattended
reboot).
Checklist: Passo 9 done; fixed the stale global tracking table (Passi 6–9
were still [ ] despite being complete). Phase B is now fully closed (1–9 [x]).
§10: single-job Measure-CIBenchmark on LinuxBuild2404 (DHCP, 10 iter) on the
Linux host — boot total 8.82s avg, near-deterministic even on DHCP (IP-acquire
7.45s, σ≈0.01). Guest-vs-guest table: Linux guest ~63% faster boot total than
the Windows static-IP guest (§8) on the same host.
plans/benchmark-windows-host.md: dual-boot plan to collect the symmetric
Windows-host numbers (§11 Win static, §12 Linux DHCP, §13 concurrent burn-in)
so the host×guest×IP-mode matrix is complete and Linux-vs-Windows is a
column-for-column comparison. Includes Windows paths, Credential Manager
prereqs, and the BaseClean-Linux snapshot caveat.
Concurrent capacity burn-in run on the Linux host as ci-runner with the
static-IP pool: WinBuild2025 40/40 PASS (~78.6s/round) and LinuxBuild2404
40/40 PASS (~70.2s/round) — 80/80 jobs, 20/20 rounds, near-deterministic
(static IP eliminates the DHCP IP-acquire variance of the §6/§7 baselines).
- RUNBOOK §9: concurrent burn-in result + ip-pool reset/recovery note.
- PhaseB-user-checklist Passo 7/8 marked done; fixed two checklist bugs:
run as ci-runner (the simone path needs g+w on the /var/lib/ci root for
ip-pool.tmp, not just the sub-dirs), and the Linux command was missing
-SnapshotName BaseClean-Linux (default BaseClean is Windows-only).
- idea-2-linux-host §8 DoD: Phase B complete (B1–B7 + >2 weeks stability).
- TODO §2.8 [P2]: ip-pool is not reconciled against live clones; a killed
job leaks its lease permanently and exhausts the 4-IP pool. Proposed fix.
Postmortem of the Linux template deploy/prepare 'no IP' investigation:
separates the three distinct causes (apt index stale from --skip-update,
deploy --force snapshot-chain, wedged host vmnet-dhcpd) and documents the
recovery (restart vmware-networks). Changelog entry summarising the
toolchain, xvfb, --force cleanup and cloud-init seal changes.
Add an --xvfb switch to 'build run' and 'job' that wraps the Linux build
command in 'xvfb-run -a sh -c ...', giving GUI toolkits (GTK4/PyGObject)
a headless $DISPLAY. Exported env vars precede xvfb-run so they
propagate into the virtual-display session. No-op (with a notice) for
Windows guests.
The local-ci-build composite action exposes it as the 'xvfb' input
(default false), forwarded as INPUT_XVFB -> --xvfb, mirroring the
use-shared-cache/skip-artifact wiring.
Docs: WORKFLOW-AUTHORING.md parameter table + Linux example note;
LINUX-TEMPLATE-SETUP.md Step 4a (the template ships python3-gi
gir1.2-gtk-4.0 xvfb). Adds build-run unit tests (wrap, plain, Windows
no-op).
deploy-linux --force reused the VM directory but left any prior snapshot
chain in place, so vmware-vdiskmanager refused the Step 4 resize with
'disk is part of a snapshot chain'. Add _clean_force_artifacts to remove
snapshot deltas (.vmdk), .vmsn, .vmsd and stale .lck before resize.
prepare-linux Step 7d seals the image: drop the stale cloud-init netplan
(leaving only the MAC-agnostic 99-ci-dhcp-allnics), regenerate networkd
config, and disable cloud-init so it never re-runs on linked clones.
deploy user-data sets package_update/upgrade: false so cloud-init does
not hold the dpkg lock at first boot and delay open-vm-tools past the
IP-detection poll; the step-7 IP timeout is raised 300s -> 600s as a
safety net for slow first boots.
Adds unit tests for _clean_force_artifacts (snapshot chain, lock dir,
clean-dir no-op) and updates prepare-linux stubs for the seal step.
Step 4a installs python3-gi gir1.2-gtk-4.0 xvfb for headless PyGObject
GUI tests, with import-gi and Xvfb assertions.
Step 1 now always runs apt-get update; --skip-update only skips the
(slow) full upgrade. Previously --skip-update skipped the index refresh
too, so a fresh package (gir1.2-gtk-4.0/xvfb) could not be located.
Linux clones could boot with the RTC interpreted as local time (guest
runs hours ahead of UTC). NTP then steps the clock backward at runtime,
landing after source mtimes were stamped, so make reported "Clock skew
detected" and future-dated files.
- build run: enable NTP and block until NTPSynchronized before stamping
the sources, so the backward correction happens pre-touch. Drop
`hwclock -s`, which read the local-time RTC and pushed the clock further
ahead.
- template prepare-linux: set-local-rtc 0 so clones interpret the RTC as
UTC and boot at the correct time; NTP then only nudges, never steps 2h.
Linux/SSH path only; WinRM branch untouched.
- Introduced a new test to verify behavior when SSH transport connects but is not ready, ensuring it returns False after the deadline.
- Added comprehensive tests for the `deploy-linux` template, covering various scenarios including missing public keys, VM directory existence, SHA256 mismatches, and failures during ISO creation and VM conversion.
- Implemented tests for SSH-related functions, including connection retries and streaming command execution, ensuring proper handling of exit statuses and progress indications.
- Enhanced existing tests to cover edge cases and ensure robustness in the deployment process.
- Implement tests for disk alert JSON output to include a message field when below threshold.
- Add tests for runner status when Gitea is online but no webhook URL is provided.
- Introduce tests for job report details to ensure no error line appears when there are no failure events.
- Add tests to skip empty JSONL files in job report listing.
- Install SIGTERM handler that raises KeyboardInterrupt so act_runner
workflow cancellations trigger the existing VM cleanup path
- Resolve guest_os before the ip_pool block and skip guestinfo IP
injection for Linux (template doesn't read VMX guestinfo to configure
network, causing SSH probe to fail against the wrong IP)
- Add -StaticIP/-Netmask/-Gateway params; injects guestinfo into cloned
VMX before start, mirroring job.py _inject_guestinfo_ip behaviour
- Add -GuestInfoOnly switch to Get-GuestIPAddress: skips vmrun
getGuestIPAddress fallback to prevent race where transient DHCP address
is returned before ci-static-ip.ps1 finishes reconfiguring the NIC
- Benchmark passes -GuestInfoOnly automatically when -StaticIP is set
- staticIP field added to benchmark.jsonl for DHCP vs static comparison
- WinRM MaxProcessesPerShell: apply 100 if below threshold before asserting
- DesktopHeap SharedSection: apply 4096 KB registry value if below threshold before asserting
- Replace cleanmgr (hangs in WinRM Session 0, no window station) with direct
PowerShell cleanup for dumps, WER files and Recycle Bin; DISM already covers
component store and WU cache
- vcpkg git clone/checkout: wrap with EAP=Continue to suppress PS5.1
NativeCommandError on git stderr (gotcha #3 from AGENTS.md)
- template.py: filter CI_EXITCODE:N marker from displayed output
- ci-static-ip.ps1: switch from vmware-rpctool (removed in recent Tools) to
vmtoolsd --cmd; use temp-file redirect to capture output from native cmd
Cleanup now runs after the system reboots post-install (fully settled
state), not right after the WU loop. Sequence: WU loop → static IP →
validation → final reboot → WinRM ready → [cleanup] → autologin →
shutdown → snapshot.
Move disk cleanup out of the Windows Update loop:
- PS scripts: extract Invoke-DiskCleanup function, add -CleanupOnly switch
for standalone post-update cleanup pass.
- Python command: always pass -SkipCleanup:$true during WU loop; after
loop exits 0, run a separate -CleanupOnly invocation (skipped if
--skip-cleanup). Fails fast if cleanup-only returns nonzero.
- Tests: 3 new tests for cleanup-runs-after-loop, skip-cleanup suppresses
it, and cleanup-failure propagates as ClickException.
- ruff --fix: remove unused shutil import, sort imports, fix UP037/UP035.
pypsrp alters terminal tty settings during WinRM sessions. This causes
readline() at the snapshot prompt to receive \r (not \n) on Enter,
printing ^M repeatedly without ever unblocking.
Fix: save termios attributes at command start (before any WinRM work,
terminal is clean) and restore them with tcsetattr(TCSADRAIN) before
the interactive snapshot prompt. Gracefully skips save/restore when
stdin is not a tty (tests, pipes, CI).
Revert the pre-flight snapshot check approach (commit 97d9dab) — checking
at start was architecturally wrong (snapshot state may change during
provisioning). The prompt stays at the end where it belongs.
pypsrp alters terminal tty state during remote sessions. Any readline()
call after WinRM work receives \r instead of \n on Enter (terminal in
raw mode), causing the prompt to spin printing ^M and never return.
Fix: check for existing snapshot and ask the user at the very beginning
of the command, before any WinRM/pypsrp connections are opened, when the
terminal is still in canonical mode. If the user declines, abort early.
At the end, if the snapshot still exists (race-safe), delete silently
and recreate — no second prompt needed.
input() with a prompt string on some terminals (zsh + Linux Mint) echoes
^M and the prompt doesn't flush before reading, causing the prompt to
appear stuck or not accept input.
Replace with explicit sys.stdout.write()+flush() + sys.stdin.readline()
which guarantees the prompt is flushed before blocking on stdin, and
readline() + .strip() handles both \n and \r\n line endings correctly.
Tests updated to use CliRunner(input=) which populates sys.stdin inside
the runner context (monkeypatching sys.stdin.readline directly is ignored
by CliRunner which replaces sys.stdin entirely).
click.confirm() doesn't strip \r — "y\r" doesn't match "y" and the prompt
loops or fails. Replace with raw input().strip().rstrip('\r').lower() which
handles both \n and \r\n line endings.
Also add --recreate-snapshot flag to skip the interactive prompt entirely
(useful for scripted re-provisioning).
Update snapshot_exists_skip/recreate tests to monkeypatch builtins.input
instead of using CliRunner input=.
execute_ps is fully blocking with no streaming. Add _run_with_heartbeat()
which runs the WinRM call in a daemon thread and prints
" ... still running (Ns elapsed) ..." every 30 s so the operator can
distinguish a hung session from a slow Windows Update pass.
Used only for the toolchain script invocation (the slow step); all other
transport.run() calls keep the direct path.
- --store-credential: after snapshot, writes build-user credentials into
the system keyring using the two-entry scheme (service:meta + service)
that KeyringCredentialStore.get() reads. Works on Windows (Credential
Manager) and Linux (SecretService / file backend).
- --credential-target: keyring target name, default BuildVMGuest (matches
config.guest_cred_target).
- Rewrote module docstring: full prose documentation of both `template
backup` and `template prepare-win` including all 12 provisioning steps.
- 2 new tests covering --store-credential with custom and default targets.
- mypy --strict clean; coverage 90.03%.
Replaces the abandoned PSWSMan approach (New-PSSession hangs on Linux)
with a Python-native implementation using the proven pypsrp WinRmTransport.
- New `template prepare-win` command in commands/template.py: starts VM,
waits for WinRM, uploads and runs Install-CIToolchain-*.ps1, handles
Windows Update 3010 reboot loop (max 10 iterations), installs CI-StaticIp
scheduled task, runs post-setup validation, graceful shutdown, snapshot.
- Helpers: _ps_escape, _wait_tcp, _parse_exit_marker, _wait_winrm.
- 40 new tests in test_commands_template.py covering all major branches.
- Fix test_commands_job.py: patch load_config in _patch_common so tests
don't pick up live /var/lib/ci/config.toml ip_pool and fail on lock.
- Coverage gate: 90.06% (was 88.95% before fix).
Without Import-Module PSWSMan the native WSMan library isn't registered
and New-PSSession hangs/fails silently. Import guarded by $IsWindows so
Windows PS 5.1 is unaffected.
WSMan TrustedHosts is a Windows WinRM client concept — does not exist
on Linux. Guard the configure+restore blocks with $isLinuxHost so the
Prepare scripts run without elevation errors from a Linux host.
-SkipCACheck/-SkipCNCheck in sessionOptions is sufficient on Linux.
Parameter not available in PS 7 on Linux. -SkipCACheck -SkipCNCheck
sufficient for self-signed lab certs; revocation check is irrelevant
for self-signed certs regardless.
Test-NetConnection is Windows-only. Replace with cross-platform TcpClient
async-connect helper (same pattern as Measure-CIBenchmark.ps1) so the
Prepare scripts run on the Linux host under PowerShell.
Windows (Prepare-WinBuild2025/2022): read ci-static-ip.ps1 from
guest-setup/windows/ on the host, push content over WinRM, register the
CI-StaticIp scheduled task (AtStartup/SYSTEM). Adds StaticIpTask and
StaticIpScript post-setup assertions.
Linux (Install-CIToolchain-Linux2404.sh): embed ci-report-ip.sh and
ci-report-ip.service content inline via heredoc tee, replacing the broken
cp-from-relative-path approach (the script runs inside the VM where the
host-side guest-setup/ directory does not exist).
template/guest-setup/linux/ci-report-ip.sh
Updated version: checks guestinfo.ip-assignment first (static path),
falls back to DHCP detection if absent. Applies static IP before
network-online.target so SSH starts on the correct address immediately.
template/guest-setup/linux/ci-report-ip.service
Drops 'network.target' from After= — runs before network is up so the
static IP is configured before SSH/other services start.
template/guest-setup/windows/ci-static-ip.ps1
Reads guestinfo.ip-assignment via vmware-rpctool at system startup (SYSTEM
account Task Scheduler task). Disables DHCP, applies static IP+gateway,
reports back via guestinfo.ci-ip. No-op when ip-assignment is absent.
template/Install-CIToolchain-Linux2404.sh
Step 7b now copies scripts from guest-setup/linux/ instead of embedding
the old DHCP-only inline script.
Eliminates VMware-Tools IP-detect polling (observed: 39–177 s variance on
Linux host B7 benchmark). When [ip_pool] is configured the orchestrator:
1. Allocates a slot IP from a JSON-backed pool (fcntl-locked on Linux)
2. Injects guestinfo.ip-assignment / netmask / gateway into the cloned VMX
before vmrun start
3. Skips _wait_for_ip entirely — uses the pre-assigned IP immediately
4. Releases the slot in _destroy_clone (all paths: success/failure/interrupt)
New modules / changes:
src/ci_orchestrator/ip_pool.py IpSlotPool + _file_lock context manager
src/ci_orchestrator/config.py IpPoolConfig dataclass; [ip_pool] TOML
src/ci_orchestrator/commands/job.py _inject_guestinfo_ip helper; pool wiring
tests/python/test_ip_pool.py 12 unit tests (acquire/release/persist)
tests/python/test_config.py ip_pool config parsing tests
tests/python/test_commands_job.py pool path + _inject_guestinfo_ip tests
config.example.toml [ip_pool] section (commented example)
Coverage 90.02%; mypy --strict clean.
4-iteration run on Linux Mint host, WinBuild2025/BaseClean.
Clone/Start/Ready/Destroy within ±20% of Windows baseline.
IP-acquire avg 99.7 s vs 58.2 s Windows — high variance (39–177 s),
attributed to VMware Tools reporting latency, not orchestrator regression.
Measure-CIBenchmark.ps1:
- Default params now Linux-aware via $IsWindows conditional
(TemplatePath, CloneBaseDir, VmrunPath, OutputDir)
- Add -GuestOS Auto/Windows/Linux param; Auto reads VMX guestOS field
(ubuntu-* → Linux, else Windows)
- Replace Test-NetConnection (Windows-only) with Test-TcpPort helper
using raw TcpClient async connect — works on PS5.1 and PS7/Linux
- Probe SSH/22 for Linux guests, WinRM/5986 for Windows guests
- Rename winrm→ready throughout ($t, JSONL field readySec, column header)
- Add guestOS field to JSONL output
_Common.psm1 / Invoke-VmrunBounded:
- Replace taskkill /F /T with platform-aware kill:
Windows → taskkill; Linux → Process.Kill(entireProcessTree) (.NET 5+)
nsis-plugin-ns7zip is a real C++ project — wrong for lifecycle testing.
burnin-dummy simulates ~30s work and writes to dist/ with no external
deps. Documents one-time prerequisites (group, keyring, Gitea push).
Without an explicit build command, job.py runs dotnet build (which fails
on nsis-plugin-ns7zip in a fresh VM) and expects bin\CIOutput as artifact
source. With a trivial BuildCommand, artifact_source defaults to "dist"
and the lightweight command always succeeds.
Also documents one-time simone group/keyring prerequisites for running
the burn-in as simone with ci-runner group access.
All shims defaulted to F:\CI\python\venv\Scripts\python.exe when
CI_VENV_PYTHON was not set. On Linux host this caused instant failure.
Now uses $IsWindows (PS 6+ automatic var; absent on PS 5.1 = Windows)
to select /opt/ci/venv/bin/python on Linux. Windows paths unchanged.
Affects: Invoke-CIJob, Get-BuildArtifacts, Cleanup-OrphanedBuildVMs,
New-BuildVM, Remove-BuildVM, Wait-VMReady, Get-CIJobSummary,
Invoke-RemoteBuild, Watch-DiskSpace, Watch-RunnerHealth,
Set-CIGuestCredential, Test-CIGuestWinRM.
Passing $psExe as a second -ArgumentList item caused double-wrapping of
the string array, collapsing all args into one string when splatted.
Computing $IsWindows inside the worker avoids the extra param entirely.
powershell.exe does not exist on Linux — child process launch failed
instantly (ExitCode null, Elapsed 00:00). Now selects powershell.exe
on Windows and pwsh on Linux via $IsWindows automatic variable.
Data from Measure-CIBenchmark.ps1 × 4 (2026-05-17, commit 5b46bba):
avg boot 60.6s, IP-acquire dominant (20–85s, σ≈26s). Adds ±20% ranges
for B7 comparison and guidance on sample size given IP-phase variance.
Burn-in (4 jobs × 10 rounds), 7-day stability run, and RUNBOOK.md
Linux section are still open. Checklist stays active until all
acceptance criteria are met.
Moves all A/B closeouts, checklists, idea docs, and implementation plans
to plans/archived/2026-05-23/. Both phases are production-stable.
Active plans remaining in plans/:
- idea-3-powershell-removal.md (Phase C — in progress)
- idea-3-esxi-support.md (Phase D — future)
- ideas-overview.md (roadmap reference)
Relocates the graphify BFS/path query script from scripts/ (a PowerShell
folder) into a new tools/ directory for Python developer utilities.
The script is not part of the CI pipeline — it's a local query tool for
navigating graphify-out/graph.json during codebase investigation.
ns7zip matrix Win+Linux passes after the WinRM quotas + desktop-heap
fixes landed in the template scripts. Monitoring window completed
without regressions.
Parallel MSBuild fan-out (e.g. ns7zip x86/x64/x86-ansi) over pypsrp
Linux→Win fails with MSBuild MSB6003 / Win32 1816 'Not enough quota'
because two default guest limits are too tight for concurrent native
processes via WinRM:
- Plugin Microsoft.PowerShell MaxMemoryPerShellMB = 150 MB
(pypsrp uses the plugin endpoint, not the Shell endpoint that
Deploy was previously bumping)
- Non-interactive desktop heap SharedSection 3rd field = 768 KB
(Session 0 csrss; CreateProcess fails when 3+ cl.exe spawn)
Pre-migration Win→Win went through Invoke-Command on PSSession with
quotas inherited from the local process, so the limits never bit.
Changes:
- Deploy-WinBuild2025.ps1 / Deploy-WinBuild2022.ps1: also set the
plugin-level MaxMemoryPerShellMB=2048, bump MaxProcessesPerShell
25→100, patch SharedSection 3rd field 768→4096 (reboot needed for
csrss to pick up, applied by post-install shutdown).
- Install-CIToolchain-WinBuild2025/2022.ps1: add Assert-Step for
plugin quota, MaxProcessesPerShell≥100, SharedSection≥4096; fail
hard if any are under threshold.
- docs/WINDOWS-TEMPLATE-SETUP.md: new "WinRM quotas e desktop heap"
section explaining the diagnosis + one-shot patch for legacy
templates; updated Step 3 validation row + troubleshooting entry.
- AGENTS.md: error #13 with symptom, root cause, and pointer to docs.
If ci-backup-template.service is killed mid-run (SIGTERM), the Python
finally block that restarts act-runner.service is skipped. A subsequent
backup then finds the runner already inactive, _stop_runner returns
False, and the finally restart is gated off — leaving the runner down
indefinitely (incident: 21 mag 2026, runner stayed off ~80 min).
Add ExecStopPost=-/bin/systemctl start act-runner.service so systemd
guarantees the runner is restarted regardless of how the unit exits
(success, failure, signal). The `-` prefix preserves the unit's own
exit status.
- Add plans/idea-3-powershell-removal.md: Phase C goal is to eliminate
the pwsh dependency from the Linux host by porting the remaining PS
scripts (bench run, validate host, smoke run) to Python sub-commands.
- Update idea-3-esxi-support.md header: Fase C → Fase D.
- Update ideas-overview.md: new four-phase table (A/B/C/D) with
prerequisiti and criteri for Phase C.
- Rename all "Hook futuri Fase C" (ESXi) → "Hook futuri Fase D" in
implementation-plan-A-B.md; add Phase C/D entries to summary and
references section.
- Update PhaseB-user-checklist.md end note and CLAUDE.md to point
Phase C → pwsh removal, Phase D → ESXi.
- Add `retention run` command (ports Invoke-RetentionPolicy.ps1): purges
old artifact/log dirs with aggressive mode when disk space is low.
- Add `template backup` command (ports Backup-CITemplate.ps1): 7z -mx=1
compressed archives in /var/lib/ci/backups/, prunes to keep-count=3,
stops/restarts act-runner.service around the copy.
- Update ci-retention-policy.service and ci-backup-template.service to use
Python; pwsh is no longer required on the Linux host.
- Fix ci-watch-runner-health.service: pass --service-name act-runner
(Linux service name, not Windows act_runner default).
- Fix _list_orphans in vm.py: wrap is_dir() inside the OSError try block
so a stat() failure on an entry is silently skipped rather than raised.
- Mark B5 complete in PhaseB-user-checklist.md.
Simone