The mwax_beamdb2fil from mwatelescope

mwax_beamdb2fil's People

Contributors

Watchers

mwax_beamdb2fil's Issues

Bugs in behaviour when we have non-contiguous dada blocks/subobs ids

For example if mwax_u2s fails mid way through and observation, then as the next obs_id comes through, mwax_beamdb2fil does not close off the old fil files properly and then fails to be write the new obs filterbank files. E.g.

[2021-11-17-19:00:48] dada_dbfil_close(): Waiting for next ringbuffer block... (Marker=32, Duration=120).
[2021-11-17-19:00:48] dada_dbfil_close(): completed
[2021-11-17-19:00:48] 327680000 bytes transferred in 4.433269s (70.4897 MB/s)
[2021-11-17-19:00:48] main: dada_hdu_unlock_read()
[2021-11-17-19:00:48] main: dada_hdu_lock_read()
[2021-11-17-19:00:48] main: dada_client_read()
[2021-11-19-10:27:47] dada_dbfil_open(): extracting params from dada header
[2021-11-19-10:27:47] dada_dbfil_open(): MODE == MWAX_CORRELATOR
[2021-11-19-10:27:47] dada_dbfil_open(): New OBS_ID detected. Starting 1321324064...
[2021-11-19-10:27:47] dada_dbfil_open(): Closing 4267111...
[2021-11-19-10:27:47] dada_dbfil_open(): Closing /datax2/mwa_fil//1321182016_20211117105958_ch76_01.fil...
[2021-11-19-10:27:47] dada_dbfil_open(): Closing /datax2/mwa_fil//1321182016_20211117105958_ch76_02.fil...
[2021-11-19-10:27:47] Obs Id:                     1321324064
[2021-11-19-10:27:47] Subobs Id:                  1321324064

The text "Closing 4267111" is telling, since that should be a filename not numbers! Something is getting confused!

Further down, we see the effect of this confusion:

[2021-11-19-10:27:48] create_fil(): filheader.nifs         : 1 Number of pols?
[2021-11-19-10:27:48] create_fil(): filheader.nbeams       : 1 Number of beams
[2021-11-19-10:27:48] create_fil(): filheader.ibeam        : 1 Beam number in this file
ERROR : error in code did not write keyword name HEADER_START correctly (written 0 + 0 bytes, expected 13 bytes)

And a complete exit of the program.

Recommend Projects