Hej MZmine team, It appears I am having a similar problem like David Lyon (posted on 2011-12-14).
I have run some environmental samples on an LTQ Velos in full scan, data-dependent MSn and PDA. The size of the raw files is circa 180 MB in the negative mode and 40 MB in positive ionization. I am working with MZmine 2.4 on an octacore server with 16 GB RAM running Ubuntu (64Bit). My heap size is 14 GB. Processing of the data according to the workflow works fine until the RAM gets full. The fuller it is the slower the processing. This appears when I run the mass detection (exact mass) even with only 16 samples. Chromatogram detection is even worse. It takes app. 6 h computation with the local minimum algorithm. Sometimes the computer manages the calculation (when I wait long enough), sometimes it produces an "outofmemoryerror: gc overhead" error. I have tried the following without any improvement: - calculate on a Windows 7 machine and on MacOS X - calculate with MZmine 2.3 - increase noise level during mass detection - crop the PDA and MSn data from the raw files (files don't get really smaller) - convert to mzXML (files get much smaller, from 180 MB to 30 MB, but same problem) I did similar calculations with similar samples in MZmine 2.2 and never experienced problems. Now I'm running out of ideas. It appears to be a problem with the JavaVM rather than with MZmine but I don't know how to solve that. Thanks for your help & all the best from snowy Germany! Martin ------------------------------------------------------------------------------ Write once. Port to many. Get the SDK and tools to simplify cross-platform app development. Create new or port existing apps to sell to consumers worldwide. Explore the Intel AppUpSM program developer opportunity. appdeveloper.intel.com/join http://p.sf.net/sfu/intel-appdev _______________________________________________ Mzmine-devel mailing list Mzmine-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mzmine-devel
