Here, you'll find a selection of commonly encountered DAQ errors, and how to fix them.
Nevis TPC Header Out of Sync
Bad NTB Timestamp
If there is a problem with the NTB timestamps, you might see an error like this pop up:
EventBuilder7_p2 | %MSG-e EventBuilder7_p2_SharedMemoryEventManager: Early 04-May-2025 12:00:12 CDT pre-events SharedMemoryEventManager.cc:1144
Event 439 has already been marked "Incomplete" and sent to art!
EventBuilder7_p2 | %MSG-e EventBuilder7_p2_SharedMemoryEventManager: Early 04-May-2025 12:00:12 CDT pre-events SharedMemoryEventManager.cc:236
Dropping fragment with sequence id 439 and fragment id 37123 because data taking has already passed this event.
In this case the NTB is what is "pushing" the DAQ (as a DAQ expert, you can change what component pushes). Once the NTB timestamps go bad, the EVBs can't build evenets because there are no NTB timestamps.
The DAQ will run for a new minutes with some errors, including:
Missing fragments from all other subsystems
SBNDGateFilter cannot find HLT
Causality timestamp issues
Eventually, the DAQ will crash or be stopped by the auto-stop script due to missing fragments.
Some symptoms that can be checked in the EVB logs are:
Error: Missing all other fragments except the NTB fragment. Cause: Bad NTB timestamp.
Error: Cannot find HLT reported by SBNDGateFilter module. Cause: The PTB fragment is missing.
How to fix this
Reset the NTB using Phoebus (insert link to instructions for Phoebus here)
3658 %MSG-e _TRACE_: SBNDGateFilter:offbeamlightfilter@BeginModule 04-May-2025 12:18:40 CDT run: 18465 subRun: 1 event: 1062 SBNDGateFilter_module.cc:202
3659 Run 18465, subrun 1, event 1062 has zero HLT word Fragments in module, not separating by beam type!