cris_rufino

Hi Lucian, I have tried with blockSize 1024, 2048 and 4096. With these values I have better performace, this meaning that I coult make a larger report than before. But I still with my original problem. I will continue to find a solution. []'s

Follow the code below, with virtualization parameters... OK, the virtualization will consume some memory. But this is causing memory problems in my app. Is the only solution limit the size of reports? Code:int blockSize=10, minGrowSize=5, maxSize=10;JRSwapFile swapFile = new JRSwapFile("/home/report/swap",blockSize,minGrowSize);JRAbstractLRUVirtualizer virtualizer = new JRSwapFileVirtualizer(maxSize,swapFile, swapOwner);params.put(JRParameter.REPORT_VIRTUALIZER, virtualizer);JasperPrint jasperPrint = JasperFillManager.fillReport(jasperFile, params, datasource);virtualizer.setReadOnly(true);JasperExportManager.exportReportToPdfFile(jasperPrint, "/home/report/file.pdf");

Hi, I using jasper 3.0.0 with virtualization, that generate files on disk avoiding outofmemory. Great!! But for too big reports (about 100.000 pages), the memory grow up and tomcat go down. Using JRSwapFileVirtualizer: - Analisyng the memory usage with VisualVM, the JRSwapFileVirtualizer routine still have high memory consumption to manage the files genereted on disk. - I did try setting the virtualize parameters (blockSize, minGrowSize and maxSize). They help, but don't solve the problem completely. Using JRFileVirtualizer: - The problem continue, several files on disk and high memory consumption for virtualization routine. Then I would like a solution that, for big reports, the virtualization routine don't have high memory consumption. The time to generate the report can be grow up in this case, but the tomcat can't go down. I think jasper should have this solution. Thanks for any response. []'s Cris.

Solved problem ;-) Setting IS_AUTO_DETECT_CELL_TYPE=true and IS_DETECT_CELL_TYPE=false make jasper 3.0.0 behavior equal jasper before 2.0.3, when the default value of IS_AUTO_DETECT_CELL_TYPE have changed, acording to chagelog file: JasperReports 2.0.3 (2007-12-12) - the default value of the deprecated JRXlsExporterParameter.IS_AUTO_DETECT_CELL_TYPE exporter parameter is now "false"; the newer IS_DETECT_CELL_TYPE exporter parameter should be used to control the cell types in the generated XLS documents. [ ]'s Cris

Solved problem ;-) Setting IS_AUTO_DETECT_CELL_TYPE=true and IS_DETECT_CELL_TYPE=false make jasper 3.0.0 behavior equal jasper before 2.0.3, when the default value of IS_AUTO_DETECT_CELL_TYPE have changed, acording to chagelog file: JasperReports 2.0.3 (2007-12-12) - the default value of the deprecated JRXlsExporterParameter.IS_AUTO_DETECT_CELL_TYPE exporter parameter is now "false"; the newer IS_DETECT_CELL_TYPE exporter parameter should be used to control the cell types in the generated XLS documents. []'s Cris Post Edited by cris_rufino at 01/11/2010 17:04

Sorry, I still have the same problem. Was the solution set the "net.sf.jasperreports.export.xls.detect.cell.type" property in jrxml file? I see this problem only in jasper 3.0 (poi-3.0). In jasper 1.3 (poi-2.0) works well. In really, I would like to solve this problem in my java project, not in jrxml file. Thanks for any response? []'s Cris. Post Edited by cris_rufino at 01/08/2010 19:07

Peaple, I was using jasper 1.3.0 to generate a XLS report using JRXlsExporter routine. The report contains string type fields with numeric value. Here, the excell and broffice calc show the cell correctly (31, for exemple). When I migrate to jasper 3.0.0, the cell is showed with leading ' character ('31, for exemple). The IS_DETECT_CELL_TYPE property did not solve my problem. My client need the cell without the ' character to export the XLS data to your internal system, i.e., my new jasper version (3.0.0) did cause an incompatibility to my client export routine. Thanks for any help response. [ ] ' s Cris. Post Edited by cris_rufino at 01/08/2010 16:40