The launcher drops RPC sessions idle >30s (easily hit while the operator sits
in the install dialogs), but a dropped socket still reports Connected until an
I/O fails. InstallProductWorker then skipped its reopen, streamed the archive
on the fresh out-of-band connection, and died on the first progress poll —
and the install-completed handler registered launch entries without checking
e.Error, so AddApp's disconnected-guard exception crashed the whole console.
- PodInfo.EnsureConnectionAlive: probe an "open" connection with a Ping and
reconnect when it is dead; used by the install and uninstall workers.
- SiteManagement.PodInfo_InstallProductCompleted: register launch entries only
on success, and surface registration failures as the row's Install Failed
state instead of an unhandled exception.
- Regression test pins the premise + recovery against vPOD's launcher server
(stale socket reports open, Ping exposes it, reconnect restores service).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Deployment intent is now implied by the install package field: if a package is
supplied the product is installed to the selected pods after being added; if not,
it is only added to the catalog. Removes the checkbox (dlgAddProduct) and the
push-decision branch (AddProduct_Click). The register-only-without-package path is
retired with it, so the now-dead RegisterProductOnSelectedPods is removed too.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Drops the standalone right-click submenu (and its handler + field). Installing a
product from the console still registers its launch entries on completion, and the
Add New Product dialog's "push now" path still uses RegisterProductOnSelectedPods,
which is retained.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reconfigure timeout (launcher reinstall of an already-configured pod):
- The console pings each configured pod and connects on success, on a loop, and
never tore that down — so it kept hammering the pod's old IP with a stale
connection while SecureConfig tried to bring the pod up on that address, which
timed out. Only a console restart (closing all sockets) cleared it.
- DeletePod now Disconnect()s the pod's connection (closes the lingering socket)
so delete-to-reconfigure works WITHOUT a console restart.
- Reconfiguring a known pod in place now pauses that pod's ping/connect loop
(OperationInProgress) and drops its stale connection before SecureConfig, and
resumes on completion (success or failure) so it reconnects with the new key.
(StopPing kills the ping thread irreversibly, so pause via OperationInProgress.)
Note: PodInfo's static mPods registry is dead code (nothing registers), so the
real fix is the disconnect, not an unregister.
Console deployment:
- Console/install.bat: installs the console to C:\Program Files\TeslaConsole,
preps the %ProgramData%\Tesla Console data dir (Users: modify), creates a Start
Menu shortcut, and adds a per-program inbound firewall rule so it can receive
pod discovery beacons. net48 is in-box, no runtime install.
- Console/build-package.bat: publishes the console (net48 framework-dependent) to
dist\TeslaConsole\App and assembles the package with install.bat.
74 tests green; the pod-config fix is UI/network behavior, validate on a pod.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-locate the two cockpit-pod projects into a single repository:
- Console/ : TeslaConsole, the net48 WinForms operator console (decompiled
reconstruction) plus its differential + catalog test suite.
- Launcher/ : TeslaLauncher, the net6 pod-side Service + Agent rewrite.
Adds a combined TeslaSuite.sln, root README documenting the shared wire
contract (and its current duplication, the main follow-up), and a root
.gitignore. Histories were not preserved per request; this is a fresh start
from the current working state of both projects.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
CydandClaude Fable 5