Ports the legacy WWSupport/APIServer MES alarm-status endpoints onto the
Inbound API, mirroring the MesMoveIn/MesMoveOut pattern; draft for review,
not yet executed. Committed ahead of PLAN-R2-08 T13's hygiene pass.
Re-ran all 8 domain reviews at HEAD 8c888f13 against the b910f5eb baseline:
every round-1 finding source-verified (168 fixed, 0 regressions, 0 false
claims); 56 new findings (1 Critical / 4 High / 15 Medium / 36 Low),
concentrated in post-baseline code (anti-entropy resync, KPI rollup
backfill, live alarm stream) and seams the fixes exposed.
Headliners: S&F resync predicate inversion can wipe the delivering node's
buffer (02-N1 Critical); resync snapshot exceeds the Akka remoting frame
size (02-N2); failover drill kills the one node keep-oldest can't survive
(01-N1); unbounded rollup backfill per failover (04-R1); live production
API key in untracked test.txt (08-NF1).
Adds PLAN-R2-01..08 + .tasks.json manifests and the Round-2 board,
P0 list, cross-plan mutexes, and wave order in 00-MASTER-TRACKER.
Alarm Summary is now live-cache-driven via a transient per-site central
live alarm cache (ISiteAlarmLiveCache) fed by a site-wide, alarm-only
SubscribeSite gRPC stream (seed-then-stream), with the 15s poll retained
as fallback + NotReporting authority. No persisted central alarm store
([PERM]). Updated Component-CentralUI, Component-Communication (new §6.1
+ SubscribeSite RPC), M7 design (T13 delivered note), native-alarms
[PERM] note, stillpending follow-up note, CLAUDE.md, and moved register
row #10 to Resolved (left #22 in place for the sibling branch).
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Adds SiteAlarmStreamEndToEndTests wiring the whole feature pipe with only the
gRPC HTTP/2 transport mocked: real SiteStreamManager.SubscribeSiteAlarms → real
SiteStreamGrpcServer.SubscribeSite → real StreamRelayActor → proto → wire
round-trip (ToByteArray/ParseFrom) → real SiteStreamGrpcClient.ConvertToAlarmEvent
→ real SiteAlarmAggregatorActor cache.
Closes the cross-cutting gap the per-task T1–T6 unit tests leave (each mocks its
neighbour): proves AlarmKey identity (instance, name, sourceRef) + full native
enrichment survive every boundary mapping, attributes and placeholder rows are
dropped end-to-end, the single site-wide stream carries alarms for multiple
instances (no per-instance filter), and a snapshot-seed row and a live delta for
the same native alarm — both mapped through the real pipe — collapse onto one
cache row.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Extend CommunicationOptionsValidator with eager bounds for the four T4
live-alarm-cache options (linger >= 0, reconcile > 0, seed concurrency 1..64,
subscribers-per-site >= 1). Enforce the per-site viewer cap fail-safe in
SiteAlarmLiveCacheService.Subscribe (reject excess viewers with a no-op
disposable rather than growing the list or throwing into the Blazor render
path). Surface two telemetry instruments on the existing ScadaBridgeTelemetry
meter: an active-aggregator observable gauge and a reconnect counter, wired from
the aggregator actor's PreStart/PostStop and its NodeA<->NodeB flip /
reconcile-driven reopen.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Wire the operator Alarm Summary page to the transient per-site live alarm
cache (ISiteAlarmLiveCache, T4). Live-cache-first: on site select the page
subscribes and rebuilds rows/rollup from near-real-time onChanged deltas; the
15s poll is kept untouched as the authority for NotReporting and as the
safety net whenever the cache is not live (pre-seed or degraded stream). Both
paths mutate shared state only via the Blazor dispatcher, so they never race,
and each rebuild is an idempotent snapshot.
- IAlarmSummaryService.BuildFromLiveAlarms: flattens a live AlarmStateChanged
snapshot to AlarmSummaryRows with the same deterministic instance-then-name
sort as GetSiteAlarmsAsync; NotReporting always empty on the live path.
- AlarmSummary.razor: inject ISiteAlarmLiveCache; subscribe on select,
re-subscribe on site change, unsubscribe on leave + Dispose.
- Tests: service BuildFromLiveAlarms flatten/sort/empty; page subscribe,
poll-fallback render, onChanged rebuild, unsubscribe on leave/change/dispose.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Update Component-KpiHistory.md with the KpiRollupHourly schema, the third
recorder rollup tick + one-shot backfill, dual raw/rollup purge retention,
raw-vs-rollup query routing by RollupThresholdHours, per-metric gauge-vs-rate
aggregation (KpiMetricAggregationCatalog), the four new options + bounds, and
30d/90d trend windows. Append a Delivered-2026-07-10 note to the m6 design
plan, move register row #22 from Deferred to Resolved, and note rollups in the
CLAUDE.md KPI History bullet. Docs-only; no code changed.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Adds the active-central-node, in-memory, reference-counted per-site live alarm
cache backing the operator Alarm Summary page. No persisted central alarm store
([PERM]) — no EF entity/table/migration/DbSet.
- ISiteAlarmLiveCache (new): singleton seam — Subscribe(siteId, onChanged) ->
IDisposable (ref-counted, linger stop on last-out), GetCurrentAlarms, IsLive.
- SiteAlarmAggregatorActor (new): one per site. Seed-then-stream ordering copied
from DebugStreamBridgeActor — open the site-wide alarm-only gRPC stream first,
buffer live deltas while the snapshot fan-out runs, flush with per-key dedup
(AlarmKey = InstanceUniqueName|AlarmName|SourceReference), then live pass-through
into an in-memory dict. Placeholder rows seeded from the snapshot and never
expected on the stream; a real-alarm delta (distinct key) never wipes them.
NodeA<->NodeB reconnect (retry budget + stability window), reconnect re-seeds,
periodic reconcile (authoritative clear-and-rebuild) corrects instance-set drift,
and a budget-exhausted stream self-heals on the next reconcile tick.
- SiteAlarmLiveCacheService (new): DI singleton facade — viewer reference-counting,
linger-delayed last-out stop (version + TryRemove(ref) race guards), the
snapshot fan-out seed (RequestDebugSnapshotAsync per Enabled instance, capped),
bounded start-retry self-heal on transient start failure, and the immutable
published-snapshot store the page reads. Cache mutated only on the actor thread;
viewer callbacks invoked outside the lock.
- CommunicationOptions: LiveAlarmCacheLinger (30s), LiveAlarmCacheReconcileInterval
(60s), LiveAlarmCacheSeedConcurrency (8), LiveAlarmCacheMaxSubscribersPerSite
(200). Task 6 formalizes eager validation + telemetry.
- DI registration + AkkaHostedService SetActorSystem wiring on the active central node.
- Tests: 14 actor (seed/stream ordering, dedup, native-alarm parity, placeholder
coherence, reconnect re-seed, reconcile replace, self-heal, stop) + 6 service
(shared start, linger stop, resubscribe cancels stop, idempotent dispose, unknown
site, transient-start self-heal). Code-reviewer pass: no persistence, no Critical;
both Important findings addressed.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Adds 30d (720h) and 90d (2160h) toggle buttons alongside the existing
24h/7d controls on all four KPI trend surfaces. No fetch-logic changes:
the setters already recompute fromUtc and re-fetch, and the query service
(T5) transparently routes windows > RollupThresholdHours to hourly rollups
— so these windows are what actually exercise the rollup path. bUnit tests
extend the Audit and SiteCalls trend harnesses to cover the new buttons.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Fold the whole raw-retention window of already-recorded KpiSample rows into
KpiRollupHourly once shortly after the recorder singleton starts, so long-range
(30 d/90 d) trend charts aren't blank until enough wall-clock passes after
deploy. Reuses Task 3's idempotent FoldHourlyRollupsAsync over
[TruncateToHour(now) - RetentionDays, TruncateToHour(now)); runs at most once
per actor lifetime (_backfillDone latch) and defers past / blocks the periodic
lookback fold via a dedicated _backfillInFlight guard so the two idempotent
upserts never race on overlapping rows. Best-effort: no exception escapes; logs
start, window, and elapsed on completion. No Host or options change.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
GetSeriesAsync now branches on window width: windows wider than
RollupThresholdHours (default 168h/7d) read pre-aggregated hourly rollups
via GetHourlySeriesAsync; windows at or below the threshold read raw
samples via GetRawSeriesAsync (preserving intra-minute detail on 24h/7d).
The boundary is strict greater-than, so exactly 168h stays on the raw
path. KpiSeriesBucketer.Bucket runs unchanged on whichever series (a 90d
rollup can still exceed the point ceiling). Both ctors route identically
via a shared FetchSeriesAsync helper reading _options.RollupThresholdHours,
the same options instance that already supplies DefaultMaxSeriesPoints.
GetSeriesAsync public signature unchanged.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Add a third Akka periodic timer (kpi-rollup) to KpiHistoryRecorderActor that
folds the trailing RollupLookbackHours of raw KpiSample rows into the hourly
KpiRollupHourly table via the idempotent FoldHourlyRollupsAsync upsert, with the
in-progress hour excluded (toHourUtc = current hour start, exclusive). A
_rollupInFlight guard coalesces overlapping ticks (mirrors _sampleInFlight), the
fold is best-effort (no exception escapes a tick), and a missed/failover tick
self-heals via the lookback re-fold. The daily purge now runs BOTH the raw
PurgeOlderThanAsync (RetentionDays) and PurgeRollupsOlderThanAsync
(RollupRetentionDays), isolated so one failure never skips the other.
New KpiHistoryOptions: RollupInterval (1h), RollupLookbackHours (3),
RollupRetentionDays (365), RollupThresholdHours (168, consumed by Task 5).
Validator adds bounds incl. the coherence rule RollupRetentionDays >=
RetentionDays so long-range trends have no data hole.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Add FoldHourlyRollupsAsync (in-memory grouped, per-metric gauge-last/rate-sum,
idempotent upsert on the series+hour key with null-ScopeKey equality, exclusive
upper hour bound so the in-progress hour is never folded), GetHourlySeriesAsync
(same KpiSeriesPoint contract as GetRawSeriesAsync), and PurgeRollupsOlderThanAsync
(one-hour-sliced batched DELETE mirroring PurgeOlderThanAsync). Adds 7 repo tests.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Add SiteStreamGrpcClient.SubscribeSiteAsync mirroring SubscribeAsync but for
the site-wide, alarm-only SubscribeSite RPC: builds a SiteStreamRequest, opens
the server stream, and delivers each event via a typed Action<AlarmStateChanged>
callback (this stream is alarm-only by contract, so Task 4's per-site cache
consumes an alarm delta with no downstream type test). Reuses the shared
enrichment mapping via a new internal ConvertToAlarmEvent helper that returns
null for any non-alarm event, defensively filtering anything that should never
appear on the stream. Factory unchanged - it already caches a client per
(site, endpoint). Adds focused unit tests for the alarm-only filter and the
test-only-client guard.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Adds an additive, site-wide, alarm-only gRPC stream backing the aggregated
Alarm Summary. Proto: new SiteStreamRequest { correlation_id } message + rpc
SubscribeSite(SiteStreamRequest) returns (stream SiteStreamEvent) on
SiteStreamService — purely additive, no field renumbering. Regenerated the
checked-in SiteStreamGrpc/*.cs.
Server: SubscribeInstance and the new SubscribeSite now delegate to a shared
RunSubscriptionStreamAsync helper (readiness/shutdown guards, correlation-id
validation, duplicate replacement, concurrency cap, bounded DropOldest channel,
relay actor, SiteConnectionOpened/Closed telemetry, guaranteed cleanup). The
only variation is the subscribe delegate: SubscribeSite calls
ISiteStreamSubscriber.SubscribeSiteAlarms (no per-instance filter). Added
SubscribeSiteAlarms to the ISiteStreamSubscriber contract (SiteStreamManager
already implements it from T1). StreamRelayActor reused unchanged — it already
drops IsConfiguredPlaceholder rows and maps the enriched AlarmStateUpdate.
Tests: SubscribeSite subscribes site alarms + removes on cancel, rejects unsafe
correlation ids, and relays a domain AlarmStateChanged as a proto
AlarmStateUpdate on the stream.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
New KpiRollupHourly entity is the hourly pre-aggregated series backbone folded
from raw KpiSample rows, letting long-range (30d/90d) trend reads scan ~60x fewer
rows. Same (Source, Metric, Scope, ScopeKey) series key as KpiSample plus
HourStartUtc bucket, folded Value, and MinValue/MaxValue/SampleCount fidelity
columns. UNIQUE IX_KpiRollupHourly_Series (upsert key + range-read cover, with the
default [ScopeKey] IS NOT NULL filter suppressed so Global rows participate) and
IX_KpiRollupHourly_Hour for purge. Non-partitioned, [PRIMARY], no DB-role restriction.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Adds SiteStreamManager.SubscribeSiteAlarms(IActorRef) — same broadcast-hub
wiring as Subscribe(...) but with no per-instance filter and forwarding only
AlarmStateChanged events (AttributeValueChanged dropped). Returns a subscription
id torn down via the existing Unsubscribe, symmetric with the per-instance
subscribe. Backs the Task 2 SubscribeSite gRPC server handler.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Two draft task-by-task plans for the two Scale/YAGNI deferrals whose revisit
triggers are now in view:
- #10 aggregated live alarm stream: transient per-site in-memory central live
cache seeded by the snapshot fan-out + additive SubscribeSite alarm-only gRPC
stream, pushed via the IDeploymentStatusNotifier pattern; honors the [PERM]
no-central-alarm-store rule. 8 tasks.
- #22 KPI hourly rollups: separate KpiRollupHourly table, recorder hourly fold
with failover-safe lookback re-fold, per-metric gauge-vs-rate aggregation,
range-threshold routing in the query service, longer rollup retention, one-shot
backfill, plus 30d/90d window buttons so rollups have a caller. 8 tasks.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Consolidates the reporter's poll cadence onto SqliteAuditWriterOptions instead of
the hard-coded 30 s constant, closing deferred-work register item #21 (the config-
shape cleanup its own XML doc flagged as a follow-up).
- SqliteAuditWriterOptions.BacklogPollIntervalSeconds (default 30).
- Reporter reads it via IOptions; precedence explicit-override (tests) > configured
> 30 s default; a non-positive value falls back to the default. Corrected the
stale "hard-code / tunable in a follow-up" class-doc.
- Cadence tests + register updated (row 21 -> Resolved).
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Closes the operator parity gap the deferred-work register tracked as #12: native
alarm source overrides could only be set one-at-a-time, while attribute overrides
had a CSV bulk path. Adds an all-or-nothing CSV import for native sources.
- Commons: extract the RFC-4180 line splitter into a shared CsvLineSplitter
(refactor OverrideCsvParser onto it — no behavior change, pinned by its tests);
new NativeAlarmSourceOverrideCsvParser (header SourceName,Connection,
SourceReference,Filter; blank = inherited).
- Commons: NativeAlarmSourceOverrideEntry + bulk SetInstanceNativeAlarmSource-
OverridesCommand (auto-registered via the reflection command registry).
- ManagementService: Deployer-gated handler — flattens once, validates every
source resolves + is unlocked + no duplicates up front, then upserts the whole
batch under a single SaveChanges (true all-or-nothing txn). Added to the frozen
authorization matrix; dispatch-coverage guard passes.
- CLI: `instance native-alarm-source import --instance-id --file` (parity with
`instance import-overrides`) + README.
- Tests: native parser (Commons), CLI parse/entry mapping, 3 bulk-handler tests
(happy path single-commit, locked-source reject, unresolved-source reject).
Also corrects a stale XML-doc line in ScriptRuntimeContext (WaitForAttribute
quality-gated mode is shipped, not "planned") and updates the deferred-work
register: marks #7/#13/#15/#16/#20 verified-resolved, #12 as CLI/API-shipped with
only the Central UI upload affordance still pending.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Joseph Doherty