• SCUBA2Errors

SCUBA-2 pipeline errors

The SCUBA-2 pipeline issues a number of error messages in red text, not all of which are fatal. This page lists common error messages and explains what they mean.

If you see an error that's not listed here, please file a bug report to the SCUBA-2 data reduction mailing list at scuba2dr@phas.ubc.ca.

Fatal errors

Most fatal errors occur when Starlink tasks exit with bad status (SAI__ERROR). If a SMURF task crashes it might indicate a problem with the data. Not all fatal errors cause the pipeline to crash but indicate problems with the instrument instead.

• SMURF crash

  !! SOCK__READSOCK: error on reading from socket
  ***** SMURF has crashed - aborting recipe *****

  • Cause: The SMURF task processing the timeseries data has crashed for an unknown reason and it was not possible to recover.

  • Action: Try re-running the pipeline. This is usually a transient error. Sometimes SMURF crashes at JAC because the /stardev installation is being rebuilt, sometimes it's due to a network (NFS) issue. If it persists, it is likely there is a problem with the data. Send a bug report to scuba2dr@phas.ubc.ca

• Zero flatfield solutions

  *** Zero (0) flatfield solutions for subarray s4a ***

  • Cause: The fastramp flatfield failed to yield a solution for the given subarray.
  • Action: Run or repeat a SETUP observation to check. Persistent failures likely indicate a DA or instrument problem.

• Bad PIXEL coordinates

  !! The supplied co-ordinates (4.07853877404664,0.408389882562362) convert to
  !     bad PIXEL co-ordinates, or lie outside the array bounds.
  !  Check that you are using the correct chosen system.
  !  BEAMFIT: Failed to fit to the beams.
  !! SAI__ERROR: Error
  Fit with beamfit failed in an unexpected manner

  • Cause: The search for a source to fit (especially a weak one) sometimes identifies pixels on the edge of the map which causes BEAMFIT to fail.

  • Action: None. The most serious consequence is that a pointing solution will not be derived.

• No good samples

  !! smf_iteratemap: No good samples
  !! SMF__INSMP: Insufficient samples for statistics
  All data flagged bad: insufficient samples available for calculating statistics - will not continue

  • Cause: MAKEMAP flagged all of the data as bad, so no output map was created. This can cause downstream problems depending on the observation. It is usually indicative of problems with the data, except in the case of closed-dome testing.

  • Action: It may be necessary to re-start the pipeline.

• Invalid Group identifier

  !! GRP1_IMPID: GRP identifier invalid; its value is 0 (possible programming
  !     error).
  !  GRP_LIST: Unable to produce a list file containing the names in a group
  !! GRP__INVID: Invalid Group Identifier given

  • Cause: Usually a data acquisition (DA) fault, whereby the same FITS headers end up being written to successive subscans. This error is only seen when running the online pipelines (QL and SUMMIT).
  • Action: None usually required. While there will undoubtedly be downstream errors, the pipeline should abort processing the current recipe and move on to the next data.

• File not found (1/2)

  !! Error accessing file
  !     '/export/data/reduced/scuba2_450/20120509/s4c20120509_00050_0003.sdf' -
  !     No such file or directory
  !  HDS_OPEN: Error opening an HDS container file.
  !  NDF_OPEN: Error opening an NDF data structure.
  !  Unable to get an NDF identifier for 's4c20120509_00050_0003'
  !! DAT__FILNF: File not found

  • Cause: The given file does not exist on disk. This errors are usually a result of NFS or disk problems. NFS issues are usually temporary.
  • Action: None may be necessary, but the pipeline will need to be re-started if the error below occurs.

• File not found (2/2)

  RECIPE ERROR: Error examining NDF children of file s4d20120509_00050_0003: Error accessing file '/export/data/reduced/scuba2_450/20120509/s4d20120509_00050_0003.sdf' - No such file or directory

  • Cause: The given file does not exist on disk. This error (which shows up in blue as it's from the perl process) causes the pipeline to crash.
  • Action: Restart the pipeline.

• Read-only file system (1/3)

  !! Error opening file
  !     /jcmtdata/reduced/dr1/scuba2_450/20120417/adam_21844/GLOBAL.sdf for
  !     read/write access - Read-only file system
  !  HDS_OPEN: Error opening an HDS container file.
  !  SUBPAR: Failed to update GLOBAL file for parameter OUT
  !  Unable to write bytes 1537:2048 to the file
  !     /jcmtdata/reduced/dr1/scuba2_450/20120417/adam_21844/cupid_mon.sdf -
  !     Read-only file system
  !  HDS_FREE: Error releasing locks on an HDS container file.
  !! DAT__FILPR: File protected
  Error in obeyw to monolith cupid_mon (task=findback): 147358683
  Arguments were: in=s20120417_00031_0004_450_fmos out=s20120417_00031_0004_450_back box=11 sub=true accept

  • Cause: The underlying file system has been marked as "read only". This error indicates a serious disk problem. The text above may differ, depending on the exact point in the pipeline that the file system state changed.
  • Action: Exit the pipeline. Fixing this error requires sysadmin privileges.

• Read-only file system (2/3)

  !! Unable to create file
  !     /export/data/agibb/data/reduced/summit_test/20120308/s8d20120308_00008_00
  !     13_ff.sdf - Read-only file system
  !  HDS_NEW: Error creating a new HDS container file.
  !  NDF_OPEN: Error opening an NDF data structure.
  !  Unable to get an NDF identifier for 's8d20120308_00008_0013_ff'.
  !  Unable to flatfield data from file 4 of 4
  !  Unable to flush written data to the file
  !     /export/data/agibb/data/reduced/summit_test/20120308/adam_16235/smurf_mon
  !     .sdf - Read-only file system
  !  HDS_FREE: Error releasing locks on an HDS container file.
  !! DAT__FILPR: File protected

  • Cause: As above.
  • Action: As above.

• Read-only file system (3/3)

  Error: Error creating symlink from ORAC_DATA_OUT to '/jcmtdata/raw/scuba2/s8a/20120308/00008/s8a20120308_00008_0014.sdf'
  Error: Read-only file system

  • Cause: As above.
  • Action: As above.

Non-fatal errors

Non-fatal errors occur when the pipeline traps certain conditions that do not prevent subsequent processing. These include known exit statuses for common tasks which can be used to diagnose problems. Some of these errors are reported for information purposes only.

• Source too weak to locate

  Unable to determine a FCF for this observation: source too weak to locate (S/N = 3.4)

  • Cause: Source is too weak to locate. The criterion is a signal-to-noise ratio of at least 10.
  • Action: None required - this message is for information purposes only.

• Source has negative flux

  Calibration returned a negative flux measurement
  Unable to determine a FCF of type ARCSEC for this observation

  • Cause: The source is not present in the map, and aperture photometry returns a negative result (due to the image not being completely flat). Usually accompanied by a similar error for the BEAMEQUIV FCF.

  • Action: None. This message is for information only.

• Decrease in number of bolometers

  Number of good bolometers for s8b has dropped by >20 %

  • Cause: The most recent fast-ramp flatfield has yielded more than 20 % fewer bolometers than the previous one.
  • Action: None usually required, though a repeated SETUP often helps.

• BEAMFIT fails to fit source (1/2)

  !! Failed to invert the curvature matrix.
  !  BEAMFIT: Failed to fit to the beams.
  !! PDA__FICMX: Failed to invert the curvature matrix to derive fit errors

  • Cause: The KAPPA BEAMFIT task was unable to fit the source. It may be too weak, or it may be at the edge of the map.

  • Action: None usually required, unless the source was expected to be bright and compact enough to fit.

• BEAMFIT fails to fit source (2/2)

  !! The returned fitted function from the Levenberg-Marquardt minimisation  is
  !     orthogonal to the columns of the Jacobian.  Try enlarging the fitting
  !     area.
  !  BEAMFIT: Failed to fit to the beams.
  !! KAP__LMFOJ: Levenberg-Marquardt fit function orthogonal Jacobian's columns

  • Cause: BEAMFIT was unable to fit the source. See above.

  • Action: As above.

• HISTAT cannot process distribution of values (1/2)

  !! There are too few values to compute ordered statistics.
  !  HISTAT: Error computing simple ordered statistics for an NDF's pixels.
  !! KPG__HSTIV: Insufficient values to compute ordered statistics
  Data may be too sparse: using stats but MODE will be undefined

  • Cause: HISTAT could not calculate the statistics for the input data (usually bolometer responsivities or noise properties), probably because a group of outliers make the distribution look bimodal.

  • Action: None - the pipeline carries on and uses stats for statistical analysis. However, the MODE will not be defined (or "NaN"), although this is rarely a problem.

• HISTAT cannot process distribution of values (2/2)

  !! Unable to determine a quantile because the data are too sparse.
  !  HISTAT: Error computing simple ordered statistics for an NDF's pixels.
  !! KPG__SPARS: Data distribution is too sparse.
  Data may be too sparse: using stats but MODE will be undefined

  • Cause: As above.
  • Action: As above.