Understanding and Solving Memory Overflow in Apache POI Excel Processing with EasyExcel

Apache POI is a popular Java library for processing Excel files, but it often causes OutOfMemory errors because it loads the entire .xlsx package (a zip of XML files) into memory.

The article explains the internal structure of .xlsx files, shows how POI's XSSFWorkbook and OPCPackage.open methods read the whole zip, and provides a code snippet illustrating the problem.

import org.apache.poi.ss.usermodel.*;
import org.apache.poi.xssf.usermodel.XSSFWorkbook;
import java.io.File;
import java.io.FileInputStream;
import java.io.IOException;

public class ExcelReadTest {
    public static void main(String[] args) {
        String filename = "example.xlsx";
        try (FileInputStream fileInputStream = new FileInputStream(new File(filename))) {
            Workbook workbook = new XSSFWorkbook(fileInputStream);
            Sheet sheet = workbook.getSheetAt(0);
            for (Row row : sheet) {
                for (Cell cell : row) {
                    switch (cell.getCellType()) {
                        case STRING:
                            System.out.print(cell.getStringCellValue() + "\t");
                            break;
                        case NUMERIC:
                            if (DateUtil.isCellDateFormatted(cell)) {
                                System.out.print(cell.getDateCellValue() + "\t");
                            } else {
                                System.out.print(cell.getNumericCellValue() + "\t");
                            }
                            break;
                        case BOOLEAN:
                            System.out.print(cell.getBooleanCellValue() + "\t");
                            break;
                        case FORMULA:
                            System.out.print(cell.getCellFormula() + "\t");
                            break;
                        default:
                            System.out.print(" ");
                    }
                }
                System.out.println();
            }
        } catch (IOException e) {
            e.printStackTrace();
        } catch (InterruptedException e) {
            throw new RuntimeException(e);
        }
    }
}

It then discusses that SXSSFWorkbook only helps with writing, not reading, and recommends using EasyExcel for reading large Excel files, which streams data to temporary files and dramatically reduces memory consumption.

package excel.read;

import com.alibaba.excel.EasyExcel;
import com.alibaba.excel.context.AnalysisContext;
import com.alibaba.excel.read.listener.ReadListener;

public class EasyExcelReadTest {
    public static void main(String[] args) {
        String filename = "example.xlsx";
        EasyExcel.read(filename, new PrintDataListener()).sheet().doRead();
    }
}

class PrintDataListener implements ReadListener
{
    @Override
    public void invoke(Object data, AnalysisContext context) {
        System.out.println(data);
    }
    @Override
    public void doAfterAllAnalysed(AnalysisContext context) {
        // after all data processed
    }
    @Override
    public void onException(Exception exception, AnalysisContext context) throws Exception {
        // handle exception
    }
}

Memory usage comparisons using Arthas are presented, showing POI consuming over 1 GB while EasyExcel stays below 100 MB for the same 27 MB file.

These findings demonstrate that EasyExcel provides a memory‑efficient alternative for reading large Excel files in Java applications.