Task #148 — Modbus block-coalescing: auto-recover from protected register holes

Pre-#148 behaviour: a coalesced FC03/FC04 read that crossed a write-only or
PLC-fault register marked every member tag Bad until the operator manually
flagged the offending tag with CoalesceProhibited. Healthy tags around the
hole stayed broken indefinitely.

Post-#148: two-stage recovery, no operator intervention needed.

1. Same-scan fallback: when a coalesced read fails with a Modbus exception
   (IllegalDataAddress, SlaveDeviceFailure, etc.), the planner does NOT
   mark members handled. The per-tag fallback in the same scan reads each
   member individually — non-protected members surface Good values
   immediately, and only the actual protected register stays Bad.

2. Cross-scan prohibition: the failed range (Unit, Region, Start, End) is
   recorded in a per-driver `_autoProhibited` set. On subsequent scans the
   planner checks each candidate merge against the set and refuses to
   re-form any block that overlaps a known-bad range. Net effect: after one
   scan with a failure, the protected range goes "per-tag mode" indefinitely
   while ranges around it keep coalescing normally.

Communication failures (timeouts, socket drops) are NOT auto-prohibited —
they're transport-level, not structural. The same coalesced read can succeed
once the transport recovers; recording it as "permanently bad" would defeat
coalescing for the whole driver instance.

Auto-prohibition state lives for the driver lifetime and clears on
ReinitializeAsync (operator restart). A periodic re-probe is a follow-up if
deployments need it without a restart.

Implementation:
- Added `_autoProhibited` HashSet<(byte, ModbusRegion, ushort, ushort)> +
  `_autoProhibitedLock` on ModbusDriver.
- `RangeIsAutoProhibited(unit, region, start, end)` overlap check called
  from the planner when forming blocks.
- `RecordAutoProhibition(...)` called from the catch (ModbusException)
  branch.
- The catch (Exception) branch (non-Modbus failures) keeps the pre-#148
  "mark all Bad in this scan, don't auto-prohibit" behaviour.
- Internal `AutoProhibitedRangeCount` accessor for tests.

Tests (3 new ModbusCoalescingAutoRecoveryTests):
- First_Failure_Falls_Back_To_PerTag_Same_Scan — three tags around a
  protected register at 102: T100 + T104 surface Good values via the
  per-tag fallback in the SAME scan; T102 surfaces the exception.
- Second_Scan_Skips_Coalesced_Read_Of_Prohibited_Range — confirms scan 2
  doesn't re-attempt the failed merge (no FC03 with quantity > 1 at the
  prohibited start).
- Tags_Outside_Prohibited_Range_Still_Coalesce — separate cluster at HR
  200..202 keeps coalescing normally even after the 100..104 cluster is
  prohibited.

234/234 unit tests green.

Follow-ups intentionally NOT shipped (smaller, independent changes):
- Bisection-style range narrowing — currently the prohibition range is the
  full failed block; the planner doesn't try to find the exact protected
  register. Operator-visible diagnostic + prohibition stays correct.
- Periodic re-probe to clear stale prohibitions.
- Surface auto-prohibited ranges through GetHostStatuses or a new
  diagnostic so the Admin UI can show what's been auto-isolated.
This commit is contained in:
Joseph Doherty
2026-04-25 01:01:42 -04:00
parent 0b7653d3b2
commit 3b0e093002
2 changed files with 193 additions and 3 deletions

View File

@@ -386,6 +386,43 @@ public sealed class ModbusDriver
/// <summary>Resolve the UnitId for a tag — per-tag override (#142) or driver-level fallback.</summary>
private byte ResolveUnitId(ModbusTagDefinition tag) => tag.UnitId ?? _options.UnitId;
/// <summary>
/// #148 — runtime-discovered ranges where coalesced reads have failed (typically because
/// the PLC has a write-only or protected register mid-block). Subsequent scans skip
/// coalescing across these ranges and let the per-tag fallback handle the members.
/// Cleared by ReinitializeAsync (operator restart) or by an explicit re-probe API
/// (not yet shipped).
/// </summary>
private readonly HashSet<(byte Unit, ModbusRegion Region, ushort Start, ushort End)> _autoProhibited = new();
private readonly object _autoProhibitedLock = new();
private bool RangeIsAutoProhibited(byte unit, ModbusRegion region, ushort start, ushort end)
{
lock (_autoProhibitedLock)
{
foreach (var p in _autoProhibited)
{
// A candidate (start..end) range is prohibited if it overlaps any recorded
// failure. Overlap rule: max-start ≤ min-end. We don't try to be smart about
// partial overlap — once a range fails, any superset of it is also untrusted.
if (p.Unit != unit || p.Region != region) continue;
if (Math.Max(start, p.Start) <= Math.Min(end, p.End)) return true;
}
return false;
}
}
private void RecordAutoProhibition(byte unit, ModbusRegion region, ushort start, ushort end)
{
lock (_autoProhibitedLock) _autoProhibited.Add((unit, region, start, end));
}
/// <summary>Test/diagnostic accessor — returns the current auto-prohibited range count.</summary>
internal int AutoProhibitedRangeCount
{
get { lock (_autoProhibitedLock) return _autoProhibited.Count; }
}
/// <summary>
/// #143 block-read coalescing planner. Groups eligible tags by (UnitId, Region), sorts
/// by start address, and merges adjacent / near-adjacent (gap ≤ MaxReadGap) into single
@@ -438,7 +475,10 @@ public sealed class ModbusDriver
var gap = tagStart - last.End - 1;
var newEnd = Math.Max(tagEnd, last.End);
var newSpan = newEnd - last.Start + 1;
if (gap <= _options.MaxReadGap && newSpan <= cap)
// #148 — skip merges that would re-attempt a known-bad range. The
// per-tag fallback will read each member individually instead.
var crossesProhibition = RangeIsAutoProhibited(group.Key.Unit, group.Key.Region, last.Start, (ushort)newEnd);
if (gap <= _options.MaxReadGap && newSpan <= cap && !crossesProhibition)
{
last.Members.Add((idx, tag));
blocks[^1] = (last.Start, (ushort)newEnd, last.Members);
@@ -477,16 +517,30 @@ public sealed class ModbusDriver
}
catch (ModbusException mex)
{
// #148 — record the failed range so the planner stops re-coalescing across
// it on subsequent scans. Per-tag fallback reads each member individually
// next time, so healthy tags around the protected hole keep working without
// operator intervention.
RecordAutoProhibition(group.Key.Unit, group.Key.Region, block.Start, block.End);
var status = MapModbusExceptionToStatus(mex.ExceptionCode);
foreach (var (idx, _) in block.Members)
{
results[idx] = new DataValueSnapshot(null, status, null, timestamp);
handled.Add(idx);
// Don't mark members handled — leave them for the per-tag fallback in
// the same scan so single-register reads can succeed for any non-
// protected member. (Pre-#148 behaviour was to mark all Bad and skip.)
// Members that ARE the protected register will fail again at single-tag
// granularity and surface the per-tag exception code naturally.
}
_health = new DriverHealth(DriverState.Degraded, _health.LastSuccessfulRead, mex.Message);
}
catch (Exception ex)
{
// Communication failures (timeout, socket drop) aren't a structural reason
// to prohibit the range — the same coalesced read might succeed once the
// transport recovers. Mark members Bad for this scan but don't auto-prohibit
// and don't deflect to per-tag fallback (which would just hit the same dead
// socket).
foreach (var (idx, _) in block.Members)
{
results[idx] = new DataValueSnapshot(null, StatusBadCommunicationError, null, timestamp);