Open mkuiack opened 5 years ago
Thanks to @gijzelaerr for pointing out this is likely due to a difference the image data protocol expected by trap, and provided by the imager.
https://github.com/transientskp/tkp/blob/137f3e8f38549e9c8eae0735c44c98b639e1e3be/tkp/stream.py#L81
I'm going to set this as a provisional target for release 6.0
This is related to the streaming goals for R7, so I move it to that milestone for now. We need to assess if the AARTFAAC streaming would be the same for the LOFAR2.0 streaming and if any of the existing AARTFAAC streaming code is useful.
I'm testing streaming trap 4.0 with two imagers streaming to localhost:9000 and localhost:9001.
Once the imagers establish the connection, TraP first times out while trying to read data. This is expected because it takes a little while for the data stream to start, after the connections are all set up between correlator, calibrators, and imager. But once the imagers start sending data TraP prints the error
error reading data: 2314926402535508307!=5136718571548659023
then TraP kills the imagers. Any ideas what this error means?Imager log:
I also saved the images to disk with the imager and verified they're all good.
mark
trap.debug.log showing attempted connections, connections, and error: