search

JStraq / PXC

Experiment Control Code
0 stars 1 forks source link

Timing errors, lost packets, general sadness #17

Open JStraq opened 4 years ago

JStraq commented 4 years ago

First run seems to go well, timing looks good, data file looks good, plotting looks good.

Second run, stuff gets weird. Instrument process: CMeas polls at proper freq, but goes for too long FileProc: somehow loses a bunch of records intermittently, but still records the correct amount

JStraq commented 4 years ago

After the first run, the sequence is somehow ending up doubled: that is, there are to CMeasCmds in the second run's metadata. Both of them, though seem to run appropriately from InstProc, tracking their time correctly and measuring the right number of datapoints at the right rate.

Additionally, every packet pushed to the FileQ is being popped and logged by the filehandler promptly, but the database seems to only write certain packets to the file.