Runtime errors and sporadic behaviour with Mopra MOPS broadband data reduction

[Kate.Brooks@…]

Hi Malte,

One of the Mopra observers is having problems with ASAP. He has MOPS broadband data and running a reduction scrip. Sometimes the script works and other times not, even on the same file. The error is not consistent. I have sent to you an example of the output error in an email.

The data are at: /DATA/KAPUTAR_3/mopra/MOPSarchive/2007-06-25_2310-M232.rp

The script is: reduction.py

Thanks, Kate

[Malte Marquarding]

This is a known bug. Eric knows about it.

MOPS datasets are very large and can create memory errors (this appeared in an external library sometime between Release 2.1 and 2.2)

ASAP has two modes of operation where a) everything gets stored in memory b) the data is stored (temporarily) on disk

a) is the default ( as single dish datasets used to be small) and can be changed via the ~/.asaprc file

Simply create a file ~/.asaprc

and put in the entry

scantable.storage: disk

This will change the default. This is the recommended way of reducing MOPS data.

This can leave some temp file in the current directory, which can be deleted after exiting asap.

This should probably go somewhere in a MOPRA user manual, so everyone knows about it.

[Malte Marquarding]

This is still happening under linux for file >600MB. The SRCTYPE column gets corrupted. I am still trying to reduce this to a minimal test case so I can get Ger van Diepen to fix this in the casacore table system.

[Malte Marquarding]

This is finally solved and will be available in asap-2.3. We should remove the note about the rc parameter setting - scantable.storage: disk is now optional.