rasimport: Segmentation fault (core dumped)

[j.oosthoek@…]

I receive an error running rasimport on ​http://planetary-earthserver.jacobs-university.de after yum update to the latest rasdaman community version.

I tried the following rasimport commands which fail:
cd /mnt/miraculix-earthserver/lighthouse_planetary_data_repository/pre_processed_early_service_DEMO_11-2011/May2012/Background/
rasimport -f themisday.tif -t GreyImage:GreySet -coll themisday
I get the following error: Segmentation fault (core dumped)

themisday.tif is HUGE (22GB) so I first though the size might be the reason.
But then I tried to add the following data which I had already added before:

cd /mnt/miraculix-earthserver/lighthouse_planetary_data_repository/pre_processed_early_service_DEMO_11-2011/May2012/CRISM/
rasimport -f crism_vnir.tif -t CRISMVNIRImage:CRISMVNIRSet -coll tempcrism
I get the same error.

Any idea what might be the problem here?

[herziga@…]

The most recent update is working fine for me, so no real idea. Things you could check though, are

• is GDAL working
• are MDD/set types correct
• is nm_meta correctly initialised

but probably you haven't changed anything else on your system but rasdaman. You could run rasimport in debug mode, which could put us on the right track where/when rasimport fails.

[j.oosthoek@…]

I reinstalled gdal, gdal-devel, gdal-java, gdal-python and now I receive the following rasimport error:

rasimport: symbol lookup error: rasimport: undefined symbol: _ZTI9r_GMarray

[j.oosthoek@…]

When I do rasdl -p I get the following error:

Using error text file: /usr/share/rasdaman//errtxts
rasdl: rasdaman schema and database manipulation tool, rasdaman v8.3 on base DBMS pgsql -- generated on 11.07.2012 18:29:26.
Warning/error in AdminIf Connect:
SQLSTATE: 08001 SQLCODE: -402

Error: connect unsuccessful; wrong connect string '/' ?
EDL003 rasdaman error: 830: base DBMS Error: Cannot connect to base DBMS server (invalid connect string in rasmgr config file?).
rasdl done.

[j.oosthoek@…]

And finally WCS also doesn't work correctly anymore:
​http://planetary-earthserver.jacobs-university.de:8080/petascope?service=WCS&Request=GetCoverage&version=2.0.0&CoverageId=hrscdtm&subsetx=x(247,443)&subsety=y(5759,5955)

[j.oosthoek@…]

I'm afraid that the following error is also related:

When I try the following query at ​http://planetary-earthserver.jacobs-university.de:8080/petascope/wcps:

for c in ( crismvnir )
          return
              encode( coverage histogram
                      over     $n x(0:9)
                      values   count(
                                   c.band90 >= 0.158676 + $n*(0.309063 - 0.158676)/10 and
                                   c.band90 < 0.158676 + (1+$n)*(0.309063 - 0.158676)/10),
                      "csv" )

I get the following error:
Error: Cannot build coverage metadata !!!

[j.oosthoek@…]

UPDATE:
The WCS and WCPS errors described in the two comments above are solved.

The rasdl -p error can be fixed by using rasdl -p —connect RASBASE

Only the rasimport error remains which again says: Segmentation fault (core dumped)

[Dimitar Misev]

I think we fixed the segfault with the latest rasimport right? Will close the ticket.