Report Failing for large Data

kapadiatap

We have our own plugin to create the Datasource and we are executing the Birt Reports from a Websphere on Linux Server.
We are using BIRT version 2.5.

We get the following exception for large data:

[1/14/12 1:19:50:376 EST] 00000021 prefs W Could not lock User prefs. Unix error code 24.
[1/14/12 1:19:50:376 EST] 00000021 prefs W Couldn't flush user prefs: java.util.prefs.BackingStoreException: Couldn't get file lock.
[1/14/12 1:20:03:332 EST] 0000001e ReportEngine E An error happened while running the report. Cause:
java.lang.NullPointerException
at java.io.BufferedOutputStream.flushBuffer(BufferedOutputStream.java:76)
at java.io.BufferedOutputStream.write(BufferedOutputStream.java:89)
at org.eclipse.birt.core.util.IOUtil.writeInt(IOUtil.java:217)
at org.eclipse.birt.data.engine.executor.cache.ResultObjectUtil.writeData(ResultObjectUtil.java:341)
at org.eclipse.birt.data.engine.executor.cache.ResultObjectUtil.writeData(ResultObjectUtil.java:249)
at org.eclipse.birt.data.engine.executor.cache.disk.DataFileWriter.write(DataFileWriter.java:105)
at org.eclipse.birt.data.engine.executor.cache.disk.RowFile.writeRowsToFile(RowFile.java:140)
at org.eclipse.birt.data.engine.executor.cache.disk.RowFile.writeRows(RowFile.java:106)
at org.eclipse.birt.data.engine.executor.cache.disk.DiskDirectExport.outputResultObjects(DiskDirectExport.java:86)
at org.eclipse.birt.data.engine.executor.cache.disk.DiskDataExport.innerExportRestData(DiskDataExport.java:142)
at org.eclipse.birt.data.engine.executor.cache.disk.DiskDirectExport.exportRestDataToDisk(DiskDirectExport.java:58)
at org.eclipse.birt.data.engine.executor.cache.disk.DiskCacheResultSet.processRestResultObjects(DiskCacheResultSet.java:81)
at org.eclipse.birt.data.engine.executor.cache.disk.DiskCache.<init>(DiskCache.java:111)
at org.eclipse.birt.data.engine.executor.cache.SmartCacheHelper.populateData(SmartCacheHelper.java:351)
at org.eclipse.birt.data.engine.executor.cache.SmartCacheHelper.initInstance(SmartCacheHelper.java:283)
at org.eclipse.birt.data.engine.executor.cache.SmartCacheHelper.getResultSetCache(SmartCacheHelper.java:244)
at org.eclipse.birt.data.engine.executor.cache.SmartCache.<init>(SmartCache.java:69)
at org.eclipse.birt.data.engine.executor.transform.pass.PassUtil.populateOdiResultSet(PassUtil.java:132)
at org.eclipse.birt.data.engine.executor.transform.pass.PassUtil.pass(PassUtil.java:62)
at org.eclipse.birt.data.engine.executor.transform.pass.PassManager.doSinglePass(PassManager.java:183)
at org.eclipse.birt.data.engine.executor.transform.pass.PassManager.pass(PassManager.java:93)
at org.eclipse.birt.data.engine.executor.transform.pass.PassManager.populateResultSet(PassManager.java:74)
at org.eclipse.birt.data.engine.executor.transform.ResultSetPopulator.populateResultSet(ResultSetPopulator.java:196)
at org.eclipse.birt.data.engine.executor.transform.CachedResultSet.<init>(CachedResultSet.java:135)
at org.eclipse.birt.data.engine.executor.DataSourceQuery.execute(DataSourceQuery.java:855)
at org.eclipse.birt.data.engine.impl.PreparedOdaDSQuery$OdaDSQueryExecutor.executeOdiQuery(PreparedOdaDSQuery.java:399)
at org.eclipse.birt.data.engine.impl.QueryExecutor.execute(QueryExecutor.java:1045)
at org.eclipse.birt.data.engine.impl.ServiceForQueryResults.executeQuery(ServiceForQueryResults.java:232)
at org.eclipse.birt.data.engine.impl.QueryResults.getResultIterator(QueryResults.java:158)
at org.eclipse.birt.report.engine.data.dte.QueryResultSet.<init>(QueryResultSet.java:98)
at org.eclipse.birt.report.engine.data.dte.DteDataEngine.doExecuteQuery(DteDataEngine.java:168)
at org.eclipse.birt.report.engine.data.dte.AbstractDataEngine.execute(AbstractDataEngine.java:265)
at org.eclipse.birt.report.engine.executor.ExecutionContext.executeQuery(ExecutionContext.java:1875)


The reports without any dynamic Computed Columns, execute without any exceptions from my local Windows machine, with 1.4 million records whereas it fails from the Linux Server.
Also, note that when I run these reports from my local machine I am using the same datasource as on the Linux Server.

I also see warning messages from BIRT :
WARNING: handle type: org.eclipse.birt.report.model.api.OdaDataSetHandle
Jan 14, 2012 1:40:23 PM null null

Can you please help?

~T~

Yaytay

What is the heap usage like on the Linux and Windows boxes?

BIRT has a habit of trying to load entire recordsets into memory, and 1.4M rows is likely to use a fair chunk of heap.
I haven't ever seen such ugly errors as a result of this though.

Jim