#52 follow-up: redeploy Z28061 on DEV + QA once the other team agrees a slot #142

Closed
opened 2026-11-15 17:25:55 -05:00 by dohertj2 · 2 comments
Owner

ScadaBridge#52 was applied on 2026-11-15 (125db032, record docs/deployment/deployment-records/2026-11-15-gate-52-delmiareceiver-fix.md). The CvdReactor.ProcessRecipeDownload template body is fixed on DEV vd03 and QA vq03/vq04, and every CvdReactor instance except Z28061 was redeployed.

Z28061 was deliberately skipped on both tiers. Another team uses it, and no redeploy slot was agreed. It still runs the old body (Children["DelmiaRecv"]), so a DelmiaRecipeDownload on Z28061 still answers "Receiver ready flag not set to true" regardless of the real flag. instance diff shows it stale=True on both tiers.

To close: agree a slot with the other team, then on each tier (pass --url explicitly; on QA use the active central):

  • scadabridge instance deploy --id 1 --url <central> ..., or
  • apply.sh apply 52 without SB_SKIP. The script step is a no-op; it deploys only the stale instance.

Verify: apply.sh check 52 shows Z28061 stale=False, and a traced flow-debug run-script --instance-id 1 --script ProcessRecipeDownload --params '{}' shows no DelmiaRecv.ReadyFlag' not found warning. Only run that traced call when DelmiaReceiver.ReadyFlag is False, because a True flag takes the write path.

ScadaBridge#52 was applied on 2026-11-15 (`125db032`, record `docs/deployment/deployment-records/2026-11-15-gate-52-delmiareceiver-fix.md`). The `CvdReactor.ProcessRecipeDownload` template body is fixed on DEV vd03 and QA vq03/vq04, and every CvdReactor instance except **Z28061** was redeployed. **Z28061 was deliberately skipped on both tiers.** Another team uses it, and no redeploy slot was agreed. It still runs the old body (`Children["DelmiaRecv"]`), so a `DelmiaRecipeDownload` on Z28061 still answers "Receiver ready flag not set to true" regardless of the real flag. `instance diff` shows it `stale=True` on both tiers. **To close:** agree a slot with the other team, then on each tier (pass `--url` explicitly; on QA use the active central): - `scadabridge instance deploy --id 1 --url <central> ...`, or - `apply.sh apply 52` without `SB_SKIP`. The script step is a no-op; it deploys only the stale instance. Verify: `apply.sh check 52` shows Z28061 `stale=False`, and a traced `flow-debug run-script --instance-id 1 --script ProcessRecipeDownload --params '{}'` shows no `DelmiaRecv.ReadyFlag' not found` warning. **Only run that traced call when `DelmiaReceiver.ReadyFlag` is False**, because a True flag takes the write path.
dohertj2 added the type:followuppriority:p3 labels 2026-11-15 17:25:55 -05:00
Author
Owner

Z28061 redeploy, 2026-11-15: QA done, DEV skipped. The owner's condition was "if its idle you can do rest of #52". Record: 061432c8, docs/deployment/deployment-records/2026-11-15-gate-52-delmiareceiver-fix.md §Z28061.

QA (active central vq03): done.

  • Idle, proven twice 67 s apart: 244/244 Good. On both sides CycleRunning False, MachineCycleID 0, FurnaceTemp 25, no batch, and no recipe download (RecipeDownloadFlag and ReadyFlag False). The only change was TrunkPressure noise.
  • Deploy: instance deploy --id 1, deployment 241, done at 23:07:36Z in 0.26 s.
  • Checks after: instance diff reports not stale; 244 attributes, all Good.
  • Read-only trace: flow-debug run-script ProcessRecipeDownload, run with ReadyFlag False so nothing is written. RouteCall and ScriptRun were Delivered with no DelmiaRecv … not found warning.
  • All ten QA CvdReactor instances now run the fixed body.

DEV vd03: SKIPPED, idle can't be proven.

  • Z28061 shows only 16 of 270 attributes Good. Every idle-check attribute reads Bad with a null value: CycleRunning, MachineCycleID, FurnaceTemp and MachineBatchID on both sides, plus the side receivers' RecipeDownloadFlag.
  • The rule is that Bad quality means not proven idle, so it was not deployed. DEV Z28061 is still stale=True on the old Children["DelmiaRecv"] body.

This stays open for DEV only. To close it, get DEV Z28061's data source back to Good quality, or get its team to confirm it's idle. Then run instance deploy --id 1 on vd03 and the same read-only trace.

**Z28061 redeploy, 2026-11-15: QA done, DEV skipped.** The owner's condition was "if its idle you can do rest of #52". Record: `061432c8`, `docs/deployment/deployment-records/2026-11-15-gate-52-delmiareceiver-fix.md` §Z28061. **QA (active central vq03): done.** - **Idle, proven twice 67 s apart:** 244/244 Good. On both sides `CycleRunning` False, `MachineCycleID` 0, `FurnaceTemp` 25, no batch, and no recipe download (`RecipeDownloadFlag` and `ReadyFlag` False). The only change was `TrunkPressure` noise. - **Deploy:** `instance deploy --id 1`, deployment 241, done at 23:07:36Z in 0.26 s. - **Checks after:** `instance diff` reports not stale; 244 attributes, all Good. - **Read-only trace:** `flow-debug run-script ProcessRecipeDownload`, run with ReadyFlag False so nothing is written. RouteCall and ScriptRun were Delivered with **no `DelmiaRecv … not found` warning**. - All ten QA CvdReactor instances now run the fixed body. **DEV vd03: SKIPPED, idle can't be proven.** - Z28061 shows only 16 of 270 attributes Good. Every idle-check attribute reads Bad with a null value: `CycleRunning`, `MachineCycleID`, `FurnaceTemp` and `MachineBatchID` on both sides, plus the side receivers' `RecipeDownloadFlag`. - The rule is that Bad quality means not proven idle, so it was not deployed. DEV Z28061 is still `stale=True` on the old `Children["DelmiaRecv"]` body. **This stays open for DEV only.** To close it, get DEV Z28061's data source back to Good quality, or get its team to confirm it's idle. Then run `instance deploy --id 1` on vd03 and the same read-only trace.
Author
Owner

Closed (owner, 2026-11-15): DEV vd03 Z28061 will not come back to Good quality. The machine has moved to QA, so its DEV data source is gone for good. The QA redeploy (061432c8) was the only part that mattered, and it is done and verified. All ten QA CvdReactor instances run the DelmiaReceiver fix. The DEV Z28061 instance still has the old body and shows as stale. That is harmless, since no machine sits behind it. Redeploy or delete it if DEV config is ever tidied up.

**Closed (owner, 2026-11-15):** DEV vd03 Z28061 will not come back to Good quality. The machine has moved to QA, so its DEV data source is gone for good. The QA redeploy (`061432c8`) was the only part that mattered, and it is done and verified. All ten QA CvdReactor instances run the DelmiaReceiver fix. The DEV Z28061 instance still has the old body and shows as stale. That is harmless, since no machine sits behind it. Redeploy or delete it if DEV config is ever tidied up.
Sign in to join this conversation.