Description
ML0's StateChannelBinarySender keeps retrying the same binary hash even after GL0 has accepted and included the state channel snapshot in its global snapshots.
Evidence
- ML0 logs show repeated
[Queue] Processing N unsent binaries with growing N
- Same binary hash retried:
f7bc1f81d9569441ce954255a10dcd40a9fa234659983f661d13f621ec807598
- GL0's
/global-snapshots/latest confirms stateChannelSnapshots contains the metagraph entry (keyed by token ID)
- GL0 ordinal: 5415, ML0 ordinal: 3347
Impact
- Cosmetic (snapshots ARE landing in GL0), but the retry queue grows unbounded
- Likely ML0 not receiving/processing GL0's acknowledgment response
Environment
- Tessellation v4.0.0-rc.10
- Local cluster (3x GL0, 3x ML0, 3x DL1)
Likely Root Cause
Framework-level issue in tessellation's binary sender acknowledgment handling. To be investigated from tessellation side.
Description
ML0's
StateChannelBinarySenderkeeps retrying the same binary hash even after GL0 has accepted and included the state channel snapshot in its global snapshots.Evidence
[Queue] Processing N unsent binarieswith growing Nf7bc1f81d9569441ce954255a10dcd40a9fa234659983f661d13f621ec807598/global-snapshots/latestconfirmsstateChannelSnapshotscontains the metagraph entry (keyed by token ID)Impact
Environment
Likely Root Cause
Framework-level issue in tessellation's binary sender acknowledgment handling. To be investigated from tessellation side.