MSSink task writes all active data streams into a separate measurement set, so a number of measurement sets are written in the parallel case (and name should be chosen accordingly to avoid conflicts).

Configuration Parameters

The program requires a configuration file be provided on the command line. This section describes the valid parameters. In addition to mandatory parameters which are always required, individual tasks often have specific parameters which need to be defined only if a particular task is used. As for all tasks, parameters are taken from keys with prefix (not shown in the table below) where name is an arbitrary name assigned to this task and used in tasklist (this allows us to run the same task more than once with different parameters, i.e. writing full resolution measurement set with one name and reduced resolution into another file). The type of the task defined by should be set to MSSink.

Parameter Type Default Description
filename string None Name of the measurement set to be written. The name can have wildcards which are substituted at the run time. %d is substituted with the UT date in the “YYYY-MM-DD” format, %t is substituted with the UT time in the “HHMMSS” format, %w is substituted by the rank (worker) number, %s is substituted with the stream number (i.e. inactive ranks are not counted). Note, the result of substitution are made consistent across all ranks, i.e. it is guaranteed that the date-stamped name will be the same for all streams. In the parallel case, the date and time are taken at the moment when first record is written to the measurement set (to be able to count active ranks, which is a run time information). In the serial mode, the date and time are taken up front when the ingest pipeline is initialised with all its tasks. In the distributed mode, a number of data-dependent substitution rules are supported. For example %b is substituted by the (0-based) beam number corresponding the data stored in the given measurement set (note, execution aborts with an exception if %b is present in the string and some data stream contains more than just one beam). Similarly, %f is replaced by 0-based index of a frequency chunk (in the order of increasing frequency of the first channel), which may be handy if data are split in frequency. The text enclosed in %{ and %} is only passed through if it is not the same for all data streams. Use %% to get a single % sign. Care must be taken to ensure that the resulting file name after substitution is unique across all data streams.
stman.bucketsize int 131072 Bucket size parameter of the storage manager associated with the measurement set (i.e. buffer size)
stman.tilencorr int 4 Tiling of the visibility cube in the polarisation dimension.
stman.tilenchan int 1 Number of channels in a single cube tile (affects caching).
stman.files string separate This option controls how storage managers are mapped to files. The separate option (default) means every storage manager works with its own file. The combined option forces the storage managers to work with just one file. This can have benefits for the lustre filesystem. The hdf5 option is similar to combined, but writes corresponding data in the hdf5 format, provided casacore has been compiled with hdf5 support. The default option uses the default casacore constructor which forces the code to take these options from the resource file and in the absence of it, defaults to separate.
stman.blocksize unsigned int 4194304 Block size in bytes for single file operations (i.e. for combined and hdf5 options, see stman.files).
stman.odirect boolean false Use ODirect o/s option, if supported, for the combined mode of storage manager operation (see above). Note, its use is discouraged.
pointingtable.enable boolean false If true, pointing table will be written for each integration. Note, it contains non-standard columns to get raw azimuth, elevation and third axis position.


########################## MSSink ##############################

tasks.tasklist = [MergedSource, Merge, CalcUVWTask, FringeRotationTask, MSSink, TCPSink]

# the measurement sets will be created in the current directory and named like
tasks.MSSink.params.filename =
# pointing table will be written
tasks.MSSink.params.pointingtable.enable = true
# storage manager parameters (caching fine tuning)
tasks.MSSink.params.stman.bucketsize = 131072
tasks.MSSink.params.stman.tilenchan = 216
tasks.MSSink.params.stman.tilencorr = 4
# type of the task
tasks.MSSink.type = MSSink