java - Appending to a workbook using poi sxssfworkbook -

i need append rows sheet of workbook. using org.apache.poi.xssf.streaming.sxssfworkbook not able achieve low memory footprint. following code:

import java.io.bufferedinputstream; import java.io.fileinputstream; import java.io.filenotfoundexception; import java.io.fileoutputstream; import java.io.ioexception; import java.io.inputstream; import java.util.arraylist; import java.util.hashmap; import java.util.iterator;  import org.apache.poi.ss.usermodel.cell; import org.apache.poi.ss.usermodel.row; import org.apache.poi.ss.usermodel.sheet; import org.apache.poi.xssf.streaming.sxssfworkbook; import org.apache.poi.xssf.usermodel.xssfworkbook;  public class excelhelper {     public static void createexcelfilewithlowmemfootprint(             arraylist<hashmap<string, object>> data,             arraylist<string> fieldnames, string filename, int rownum) {         try {             if (rownum == 0) {                 // creating new workbook , writing top heading here                 sxssfworkbook workbook = new sxssfworkbook(1000);                 sheet worksheet = workbook.createsheet("sheet 1");                 int = 0;                 iterator<string> it0 = fieldnames.iterator();                 row row = worksheet.createrow(i);                 int j = 0;                 while (it0.hasnext()) {                     cell cell = row.createcell(j);                     string fieldname = it0.next();                     cell.setcellvalue(fieldname);                     j++;                 }                 rownum++;                 fileoutputstream fileout = new fileoutputstream(filename);                 workbook.write(fileout);                 fileout.flush();                 fileout.close();             }             inputstream filein = new bufferedinputstream(new fileinputstream(                     filename), 1000);             sxssfworkbook workbook = new sxssfworkbook(                     new xssfworkbook(filein), 1000);             sheet worksheet = workbook.getsheetat(0);             iterator<hashmap<string, object>> = data.iterator();             int = rownum;             while (it.hasnext()) {                 row row = worksheet.createrow(i);                 int j = 0;                 hashmap<string, object> rowcontent = it.next();                 iterator<string> it1 = fieldnames.iterator();                 while (it1.hasnext()) {                     cell cell = row.createcell(j);                     string key = it1.next();                     object o = rowcontent.get(key);                     if (o instanceof string) {                         cell.setcellvalue((string) o);                     } else if (o instanceof double) {                         cell.setcelltype(cell.cell_type_numeric);                         cell.setcellvalue((double) o);                     }                     j++;                 }                 i++;             }             filein.close();             fileoutputstream fileout = new fileoutputstream(filename);             workbook.write(fileout);             fileout.flush();             fileout.close();         } catch (filenotfoundexception e) {             e.printstacktrace();         } catch (ioexception e) {             e.printstacktrace();         }     } } 

i appending file passing content in batches(so save on jvm memory) , incrementing variable rownum.

as per understanding, when re-opening file

sxssfworkbook workbook = new sxssfworkbook(new xssfworkbook(filein),1000); 

the constructor xssworkbook reloads complete file in memory, result in gc limit exceeded.

i went through http://poi.apache.org/spreadsheet/how-to.html unable find suitable solution usecase.

can guys please suggest how fix achieve low memory footprint appending rows workbook?

sxssfworkbook doesn't need output loaded in memory management. write of data @ once. if try loading whole workbook stores in memory, when writing @ once uses storage space instead. 1000 lot put in constructor on computers. if want, try putting 100 or other lower number in constructor instead of 1000.