I restarted the DAQ to clear the 0x2000 status of h1susetmy due to INI file change (2.9 upgrade). This was not a clean DAQ restart. About 50% of the FE models went into a 0xbad DAQ status, which was cleared by restarting the mx streamers. This was complicated by an incorrect start_streamers.sh being in controls search path. The incorrect one is configured for LLO.

A second problem was that both framewriters restarted themselves 5 minutes into the new run, a problem which was not repeated.

The third problem is that h1susetmy had constant CRC errors from h1dc0.

Thinking that perhaps a mismatch of 2.8.4 and 2.9 models at EY was a possible cause of the CRC problem, I recompiled h1iopsusey and h1sustmsy against 2.9. Now both ETMY and TMSY had CRC errors.

The problem was tracked to the ODC channel in the SUS models using the new UINT32 data type. Until we get permission to upgrade the DAQ tp 2.9 we hand edited H1SUSETMY.ini and H1SUSTMSY.ini and changed the one instance of "datatype=7" to "datatype=2". Both models and DAQ were restarted and all are now happy campers.