One key per file: ScadaBridge:Cluster:SplitBrainResolverStrategy auto-down ->
leader-auto-down in appsettings.Central.json and appsettings.Site.json, build
061432c8 unchanged. G6 boot lines on both services, health 200, MxGateway and
Z28062 at baseline, bogus key 401, Overview 6/6 Up. Downtime ~23 s central,
~30 s site. No drill: single-node clusters.
Claude-Session: https://claude.ai/code/session_01GT67AFKC45MBz1wPZnVbX8
vq04 (active central) killed with SCM auto-restart disabled and held down 94 s:
UNREACHABLE +11.4 s, [leader-auto-down] DOWNING +26.4 s, survivor vq03 active
+26.7 s. Recovery config restored byte-identical before Start-Service; vq04 rejoined
as standby in ~4.7 s with no exit-75. Also corrects the earlier Drill A 'sampler gap'
(an awk keying error; vq04 was ready at +20.8 s).
Claude-Session: https://claude.ai/code/session_01GT67AFKC45MBz1wPZnVbX8
Both centrals killed in turn (Stop-Process on the service PID). SCM restart in ~5 s;
the survivor downed the old incarnation at ~+9-10 s and took the active role; the
victim rejoined as standby at ~+21 s. leader-auto-down stood aside (already Down),
so its 15 s timer path was not exercised on QA. Site-pair drill still pending
(Z28061 slot, #142).
Claude-Session: https://claude.ai/code/session_01GT67AFKC45MBz1wPZnVbX8
CvdReactor.ProcessRecipeDownload now addresses Children["DelmiaReceiver"] on DEV vd03
and QA vq03/vq04. Redeployed DEV Z28062 + Z28061Sim, QA Z28062-Z28069; Z28061 left
stale on both tiers (shared with another team, no slot agreed). Traced read-only call
on Z28062 (ReadyFlag False): the pre-fix 'DelmiaRecv.ReadyFlag not found' warning is
gone after the fix.
apply.sh: every CLI call passes --url/--username/--password explicitly (an env-only
SCADABRIDGE_URL silently reached DEV), plus SB_EXPECT (refuse unless the expected
instances exist = tier proof) and SB_SKIP (never deploy the listed instances).
Record: docs/deployment/deployment-records/2026-11-15-gate-52-delmiareceiver-fix.md
Claude-Session: https://claude.ai/code/session_01GT67AFKC45MBz1wPZnVbX8
SourceNode backfill 843 rows (0 NULL), AuditLog__HashChain__Enabled=true on both
centrals (ChainBacklog unset, head export off, single login), pair stopped together and
started vq03 first: central API ~29 s, sites not restarted. One anchor at
2026-11-13T14:07:43.3208169Z, 0 pre-cutoff rows chained, 20 new rows ChainSeq 1-20,
verify-chain 2026-11 Intact on both centrals, 2026-10 NotChained. MxGateway 2196 Good,
WarmCanary 2 Good, LocalDb sync connected. Backup ScadaBridge_pre-hashchain_20261113T090524.bak.
Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
Connection 2 / template 27 / instance 16 on site-a, bound to the OtOpcUa script-runner
canaries; serviceLevelFailover at its defaults. Under Warm the connection moved off the
Standby by ServiceLevel (connect-time preference, 0.1 s) and on the booting Standby (50);
a graceful Primary stop still took the connectivity path (66 s, #141). Kept as monitor.
Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
Pair window per pack 8ffa6394: COPY_ONLY backup + VERIFYONLY, stop both pairs,
idempotent EF script 58->64 from the Mac (flip in ~2 s), swap, centrals then
sites started together. Site data path ~2 min 6 s, central API ~58 s.
leader-auto-down started on all 4 nodes; no #116 drills; hash chain off;
recovery flag TRUE on all 4 (actions kept). Env digest-identical; only the
strategy key changed in appsettings; per-node JWT/idle values preserved.
Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
No migration (64). ~45 s site outage. SCM non-crash recovery flag set for exit 75 (#132).
#131: SBTEST01/02 AlarmStatus on ScriptAlarm.Priority, verify 13/13 before and after.
Hash chain: SourceNode backfill 843 rows, Enabled=true (ChainBacklog unset, head export off),
one anchor at 2026-11-13T12:16:53Z, 0 pre-cutoff rows chained, verify-chain 2026-11 Intact.
Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
- Target 3a59c3f6 -> 8e32f353 (adds #138 ServiceLevel failover, #139 chain
cutoff, #140 graceful stop, LocalDb 0.3.8). No migration in the range;
the idempotent 58->64 script regenerated at 8e32f353 is byte-identical
(SHA-256 488da93d...c34ecc0).
- Owner's go 2026-11-13 ("just deploy"): #116 leader-auto-down flip and the
exit-75 service recovery setting in; no #116 drills; hash chain and head
export off; #52 stays excluded.
- #131 re-confirmed DEV-only against the QA config DB.
- Preflight re-run 07:05 EST: 58 migrations, 3,441 AuditLog rows, EventId
key, log 63 MB used of 200.
Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
Read-only export of DEV vd03 and QA vq03 (2026-11-12): the only
Children["DelmiaRecv"] user is CvdReactor.ProcessRecipeDownload (script 4,
identical on both envs; DEV 3 / QA 9 instances), and the only ScriptAlarm
.Severity reader is DEV TestMesMachine.AlarmStatus (script 42, SBTEST01/02).
QA has no #131 work.
Adds the original (rollback) and fixed bodies, a guarded check/apply/rollback
script that re-sends every script field, and the apply plan: #52 any time,
#131 with or after the dd4f1c2a binary. The old #131 body still compiles on
main but returns banded values; the fixed body is what fails to compile on the
deployed binaries. Rig evidence for both gates is in the README.
Refs #52#131
Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
Pins ZB.MOM.WW.LocalDb, .Replication and .Contracts at 0.3.7 (published
2026-11-12). Wire-compatible; no config or code change needed.
- 0.3.6 (scadaproj#36): the library sync service enforces
LocalDb:Replication:ApiKey itself. LocalDbSyncAuthInterceptor is kept for
its refuse-every-stream-when-blank rule; its doc comment is updated. The
rig's site-a pair already carries one identical key on both nodes.
- 0.3.7 (scadaproj#16): a tombstone is held until the peer has it, so a
node offline past TombstoneRetention no longer resurrects deleted rows.
The survivor's ledger (sf_messages / site_events deletes) grows for the
length of the outage; optional TombstoneHardCap bounds it.
Deploy notes added to the topology + installation guides, the S&F
requirement and CLAUDE.md: stop/start the pair together, same ApiKey on
both nodes, ledger growth while a peer is down.
Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
vd03 is the DEV box (vd = DEV, vq03/vq04 = QA). Historical docs called it prod/production.
Only vd03-referential uses are changed; generic 'production' wording (production HOCON,
production code paths, ASPNETCORE_ENVIRONMENT=Production, real-world production guidance)
is unchanged. One verbatim git commit message quoted in PLAN-07 is left as recorded.
SitePskProvider cached each site's gRPC preshared key forever; only site removal
invalidated it, so rotating SB-GRPC-PSK-{site} broke the central<->site link on
both central nodes until they restarted. The cached key is now dropped when:
- the secret is written through /admin/secrets: central replaces the secrets
library's ISecretCacheInvalidator with SitePskRotationInvalidator, which also
publishes SitePskInvalidated on a new DistributedPubSub topic
(site-psk-invalidation) so the peer central node drops its copy and its
resolver cache entry too (SitePskInvalidationSubscriber);
- a site rejects a central->site call with PermissionDenied/Unauthenticated:
WithSiteCredentials wraps every per-site channel's handler in
PskRejectionRefreshHandler, so the next call re-reads the key (covers
out-of-process writes such as the secret CLI);
- a site->central call presents a token that does not match the cached key:
CentralControlAuthInterceptor invalidates and re-reads once before refusing;
- ScadaBridge:Communication:SitePsks reloads.
Invalidation is case-insensitive because SecretName lowercases names. Rotation
runbook (topology-guide), Component-Communication/Host and the #85 cache audit
(D2) updated.
Claude-Session: https://claude.ai/code/session_01GT67AEHd2Acv1YPwGfXEoV
I2: topology guide + Component-SiteRuntime + register row 127 state that a site pair using
deployed secrets must run LocalDb replication. I3: row 127 + topology guide state the upgrade
impact (whole artifact deploy fails and reads Out of date until referenced secrets exist
centrally). M6: audit row carries secrets/missing, no removed field; removal-only deploys write
no central row. M2/M3/I1 as-built notes; CLAUDE.md secrets bullet corrected.
Second vd03 window of the day: no migrations, appsettings.Site.json gains ScriptFileAccess:AllowedRoots
(pushed over SMB before the swap), sample recipes in E:\ApiInstall\ScadaBridge\recipes. The Delmia
test receiver now reads + parses the downloaded file; real HTTP R100 -> 12 values loaded, R900 -> rejected
with line numbers. verify-mes-test-machine.sh: D1's cleared check reads through BTDB, so it only runs
with group C (the rig runs VERIFY_GROUPS=AD).
vd03 (DEV) upgraded 2b9a7c3c -> fe89363b via the deployment-only winrm channel: COPY_ONLY backup,
55->57 delta script (AddAlarmMesReportable, AddSiteArtifactRevision), ~53 s window, snapshots
identical to baseline, artifacts deployed. Register row 126 now covers the QA pair only.
docker/seed-mes-test-machine.sh (idempotent) builds a standalone TestMesMachine base template -
Left/Right TestMESReceiver + TestDelmiaReceiver (prod attribute names, STATIC), computed alarms
with IsMesReportable, CvdReactor-compatible MesMoveIn/MesMoveOut/ProcessRecipeDownload/AlarmStatus
scripts ported to SetAsync+WaitAsync, and Galaxy-side simulators with reject/no-response knobs -
plus instances SBTEST01/02 and the SimpleAlarmStatusRequest (was a stub) / AlarmStatusRequest
inbound methods in the legacy /mes/(simple)alarmstatus shapes, resolving machines through BTDB.
BTDB rows SBTEST01=990001 / SBTEST02=990002 added on vd03.
docker/verify-mes-test-machine.sh drives the real inbound methods via flow-debug: PASS 25/25.
Both driven from the Mac over the winrm CLI (WinRM 5985 + SMB 445, no helper
on the boxes). Migration 53->55; the first Central boot encrypts
DataConnections.PrimaryConfiguration (dp:protocol:v1:), so rollback is the
COPY_ONLY DB restore, not a binary swap. QA: all four nodes stopped together,
one shared-DB migration, 2 min 35 s window, preflight against a restored clone.
Claude-Session: https://claude.ai/code/session_01RgthVLVU5CJ8SUuEwruTkt
Row 104: docker/vacuum-site-localdb.sh stops both nodes of a pair together, sets
auto_vacuum=INCREMENTAL + VACUUM on each node volume with integrity_check before/after,
then starts both (DRY_RUN=1 reports). Run on all three rig pairs: six files 0 -> 2, ok,
~11% smaller; site-a replication reconnected with backlog 0 and no watermark change —
VACUUM is replication-invisible (ledger keys by declared PK, oplog seq is INTEGER PRIMARY
KEY AUTOINCREMENT). Topology guide gains the recipe.
Row 105: SiteLocalDbReplicatedTableFkGuardTests builds the real site schema via OnReady
(unreplicated config so 0.3.3's own guard cannot mask the culprit) and asserts every
SiteLocalDbSetup.ReplicatedTables entry (now internal) exists and has no inbound
CASCADE/SET NULL/SET DEFAULT foreign key, with a negative control. The constraint is
enforced at build instead of at the first replicated boot.
Row 106 (found during 104's live verification, NOT caused by it — the identical 127k-row
full snapshot also landed at the earlier deploy restart): SnapshotStreamer.SendAsync stamps
as_of_seq = MAX(seq), which is 0 once a converged sender's oplog is ack-pruned to empty;
a zero-watermark receiver (rebuilt node) is never lifted, so the empty-oplog horizon rule
re-streams a full snapshot on every reconnect. Sender-side one-line fix for LocalDb 0.3.4:
as_of_seq = max(MAX(seq), last_acked_seq).
Row 79 executed its own prescribed remedy without waiting for a second
corruption: all ten rig data dirs (8 docker/ nodes + 2 docker-env2/ site
nodes) move from virtiofs bind mounts to named docker volumes
(scadabridge-<node>-data), seeded in place with both rigs stopped and
live-verified (site-a localDbReplicationConnected=true / backlog 0 over
the migrated files; env2 stable, 0 restarts). The feared inspectability
loss is moot — OrbStack exposes volumes at ~/OrbStack/docker/volumes/.
The rebuild surfaced latent env2 drift (pre-Secrets-0.6.2 image): env2
now carries the x-secrets-store-env posture + central data volumes.
Row 93's silent maintenance trap is now loud: SiteEventsSchemaCalibrationTests
pins the site_events columns/indexes the EventLogPurgeService size model is
calibrated against, failing with a pointer to the memo §4 recalibration on
any schema drift. Rows 92 (accepted by design) and 94 (standing tripwire)
annotated accordingly.
Row 81 — the cap could not be enforced on a replicated node. MaxStorageMb was
compared against page_count-freelist_count over the WHOLE consolidated LocalDb
file, which since Phase 1/2 holds ten tables plus __localdb_oplog /
__localdb_row_version / __localdb_snapshot_state. Every 1000-row trim batch freed
~390 KB of site_events pages and added ~1000 oplog rows + ~1000 tombstones to the
same file, so the measured size did not fall, the "size did not decrease" guard
fired, and the trim stopped 2.5x over a 64 MB cap with 226 k rows still present.
A node whose peer was down could not enforce its cap at all, and the event log's
budget was in practice shared with bookkeeping it does not own.
The cap now measures site_events AND ITS FOUR INDEXES only, so replication
bookkeeping can neither consume nor block the budget — and the cap stops being a
function of the peer's health, which was the wrong direction of coupling: an
inflated oplog is a replication fault, and letting it drive event deletion would
destroy operational history in response to one.
dbstat — the exact per-table measurement the register proposed — is NOT usable
here, verified before implementing. It is a compile-time option and the bundled
SQLitePCLRaw.lib.e_sqlite3 (SQLite 3.53.3) does not enable it: PRAGMA
compile_options lists no ENABLE_DBSTAT_VTAB, pragma_module_list returns ten
modules with neither dbstat nor sqlite_dbpage, and selecting from dbstat fails
with "no such table". Switching SQLite builds was rejected — the provider is
shared with ZB.MOM.WW.LocalDb and every other consumer, so it trades cap accuracy
for fleet-wide version skew. The footprint is therefore modelled from sampled row
width (500 rows each end, octet_length so multi-byte UTF-8 is not under-counted)
through a page-layout model; whole pages rather than a flat rate, because page
fill collapses once rows approach half a page and a linear model underestimates
that by ~40%. Clamped by whole-file logical size, a true upper bound that can only
correct an over-estimate and never binds on a replicated node.
Index pages count against the cap deliberately: measured at ~34% of the table's
footprint, and the 1 GB default was sized against a 393 B/row figure that already
included them.
Guard semantics restated — two backstops replace the single on-disk check:
(1) row width unsampleable => refuse to trim (checked before the cap comparison,
since a zero width makes "over cap" unanswerable rather than false; the old loop's
answer to an unanswerable cap was to keep deleting); (2) estimated size not
falling across a batch => stop. (2) is the old guard re-based onto the
table-scoped metric, so it can no longer be tripped by replication bookkeeping and
now means exactly one thing: events are arriving as fast as the trim removes them.
For that to be observable the loop re-reads the row count each batch, while
bytes-per-row is sampled once per tick and held fixed — re-sampling per batch
would jitter the estimate as the trim eats the head of the table and could stall
the loop through guard (2), the same self-defeating measurement as row 81 itself.
Row 84 — LogReplicationBacklogNote moved from per-phase to once per tick, driven
by the shared PurgeBudget. A tick where retention AND the cap trim both deleted
used to emit two identical Information notes; the tick total is also the more
useful number beside an oplog spike.
Measured model error: -0.8% / +8.4% / +12.8% across representative workloads,
biased high on purpose (an early cap forfeits budget; a low one fails at its only
job). Independently cross-checks row 34's harness at 397.2 vs 393 B/row.
Fail-before: reverting the estimator to whole-file semantics fails 5 of the new
tests and reproduces the original signature exactly (remaining=2000
estimate=21725184 cap=11534336 — rows destroyed, cap still unmet); restoring
per-phase notes fails the row-84 test with "collection contained 2 matching items".
No config keys added or renamed, so no rig/deploy changes. Decision record:
docs/plans/2026-08-17-site-event-log-cap-semantics-decision.md
Tests: SiteEventLogging 104/104 (was 93), Host 520/520, solution build clean.
Measured the unbounded purge burst and found it can force a full ten-table
snapshot resync, then bounded it. No purge-exemption path was added — the
deletes are still CDC-captured like any other write, which stays ruled out
by design; what changed is how many a single tick performs.
MEASURED (harness: real EventLogPurgeService over a real LocalDb file with
site_events genuinely RegisterReplicated, plus a two-node loopback gRPC
replication pair; docker rig untouched):
393 bytes/row incl. indexes => a 1 GB event log holds ~2,731,947 rows.
Expiring all of them in ONE tick: 285 s, 2,731,947 oplog rows,
284 MB oplog payload, 2.73 M row-version tombstones (7-day retention),
i.e. 2.7x LocalDb's MaxOplogRows default of 1,000,000. Crossing that makes
the 60 s maintenance tick prune the oplog and set needs_snapshot — a FULL
snapshot of all ten replicated tables, bought by a routine retention purge.
The resident backlog is produced-minus-drained, so a healthy peer blunts it:
on loopback the same burst peaked at 391,424 and cleared in 6.0 min. That is
the problem, not the reassurance — the old safety margin was a function of
peer throughput and vanished with the peer down. A bounded tick is safe at
any peer speed.
Drain is not a bottleneck: 7.6k-19.8k tombstones/s (20k / 500k / 2.73M runs).
SHIPPED:
- SiteEventLogOptions.MaxPurgedRowsPerTick (default 250,000; 0 = unlimited
escape hatch, negatives rejected). One SHARED budget across the retention
purge and the cap trim — two independently-capped phases would let a tick
emit twice the intended burst. 4x margin under MaxOplogRows, which is also
shared with every other replicated table's ordinary traffic.
- SiteEventLogOptions.PurgeCatchUpInterval (default 5 min, validated > 0 and
<= PurgeInterval). Without it pacing would make retention converge in 11
DAYS at the daily cadence; with it the worst case converges in ~11 ticks
(~55 min), every burst bounded. 5 min rather than 1 so ticks cannot outrun
the peer's drain and re-accumulate the backlog.
- A ceiling is claimed only when work genuinely remains (bounded EXISTS probe),
so an expired set landing exactly on the budget does not arm a no-op retry.
- Visibility: rate-limited (15 min) Warning naming the ceiling and the
accelerated interval, plus an unthrottled counter
scadabridge.site.event_log.purge_ceiling_hits — its SHAPE distinguishes a
backlog converging from a site whose event rate outruns one paced tick.
- The purge loop is a delay loop rather than a PeriodicTimer, so its cadence
now anchors to the END of the previous purge. Documented.
ROW-33 INTERPLAY VERIFIED: pacing lowers each spike and lengthens total purge
duration, but a full 250k tick drains in ~23 s (~4 min even at 10x a real
link's penalty) against the 15-minute sustained dwell and a 5-minute gap to
the next tick. The UNPACED 2.73M burst had only 2.5x of margin (6.0 min), so
pacing makes the badge assumption robust rather than merely preserving it.
Docs travel together: Component-SiteEventLogging.md, topology-guide.md
("Reading the replication backlog" + the row-33 badge note), CLAUDE.md's
purge-burst sentence, and a dated measurement/decision note under docs/plans/.
Fail-before: the new behavioural tests do not compile against the pre-fix
service (RunPurge was void); with PurgeBudget neutralised to always-unlimited,
6 of the new tests fail and all 19 pre-existing ones pass.
Tests: SiteEventLogging 93/93, Commons 715/715, HealthMonitoring 115/115,
Host 505/505. Solution build clean.
Documents the host-level hazard behind row 69's LocalDb corruption, all
live-verified 2026-08-15. Scope is the macOS docker rigs only — this is a
property of the filesystem the rig's data files sit on, not a ScadaBridge defect
and not something a production Linux deployment is exposed to.
Every rig data file (site-localdb.db, auditlog.db, secrets) rides a relative
bind mount out of the repo tree, which on macOS crosses virtiofs — and virtiofs
does not give SQLite true fsync durability. `--force-recreate`, which deploy.sh
uses, is exactly the abrupt mid-write kill that turns that into a torn file.
Records both observed failure modes, and specifically how differently they
present, which is the part that makes this hard to spot: site-a-b failed loudly
while running (every sync call throwing, connected:false, backlog 892), while
site-a-a was corrupt LATENTLY and only became fatal on the next upgrade, when a
new build's boot-time re-registration read pages the old build never touched
(SIGABRT crash-loop x12). A clean-looking node proves nothing about its file.
Second mode is the VM idle-sleeping mid-operation — `vmgr ... msg=sleep` in
~/.orbstack/log/vmgr.log, then virtiofs `refresh_nodeid: race with new lookup`
errors and a hung docker until `orb restart docker` (data intact, containers
auto-restart). Written so a future session checks vmgr.log BEFORE investigating
ScadaBridge, which is where ~20 minutes went last time.
Recovery runbook is the one that actually worked: stop the pair together,
quarantine site-localdb.db AND its -wal/-shm sidecars with a .corrupt-<date>
suffix (never delete — forensics; and a stale WAL beside a fresh database is its
own corruption), restart together, then VERIFY convergence rather than assuming
it. Notes the precondition that this is a replicated pair's recovery — it leans
on the startup reconcile plus CDC snapshot resync, so an unreplicated node has
no peer to reseed from and needs a backup instead.
Mitigation options tabulated with their real costs (named volumes = native ext4
but loses the host-side inspectability the debugging sessions lean on;
disabling OrbStack idle-pause = a host setting that cannot be version-controlled
and silently does not exist on a new machine; accept + runbook = current), so
the status quo reads as a choice rather than an oversight.
Docs only — no compose change, no host config change, rig untouched.