Java Heap Memory error

后端 未结 6 1196
遇见更好的自我
遇见更好的自我 2021-01-06 05:40

I am getting this error:

Exception in thread \"main\" java.lang.OutOfMemoryError: Java heap space
    at com.mysql.jdbc.MysqlIO.nextRowFast(MysqlIO.java:1585         


        
相关标签:
6条回答
  • 2021-01-06 06:12

    When you query returns very large sets, the mysql driver will by default load the entire result set in memory before giving you control back. It sounds like you're just running out of memory, so increase the memory furter, e.g. -Xmx1024M

    The other alternative might be to enable streaming results on your MySQL queries- this prevents the entire ResultSet to be loaded into memory all at once. This can be done by doing

    stmt = conn.createStatement(java.sql.ResultSet.TYPE_FORWARD_ONLY, 
               java.sql.ResultSet.CONCUR_READ_ONLY);//These are defaults though,I believe
    stmt.setFetchSize(Integer.MIN_VALUE);
    

    (Note that anything else but Integer.MIN_VALUE has no effect on the MySQL driver)

    0 讨论(0)
  • 2021-01-06 06:18

    You don't need to worry about -Xms too much - that's the initial Heap size on startup.

    -Xmx is the max Heap size. Try increasing it until either you don't get the OutOfMemory exception anymore, or until you run out of memory.

    If you run out of memory, you can't load the entire ResultSet at once, and you'll need to apply the approaches that others have mentioned.

    0 讨论(0)
  • 2021-01-06 06:22

    Is Lazy Load applicable? Can you use an upper limit to bound the results of a piece-wise query?

    0 讨论(0)
  • 2021-01-06 06:25

    Try splitting your data into a number of result sets. Think about what you want to do with the data once you got it back from the database. The result set is too large to fit in heap space.

    0 讨论(0)
  • 2021-01-06 06:26

    Must the whole result from the query be in memory? If yes then you should buy more RAM. If not, then partitionning the problem will be the proper solution. You can even let the database consolidate the data for you in many cases.

    If your solution to a given problem doesn't scale well with the amount of data present you will return to this problem even by extending your RAM. So the really good solution is to avoid the situation at all.

    0 讨论(0)
  • 2021-01-06 06:28

    I discovered that the MySql driver has a memory leak. I use it with tomcat in a connection pool setup. Loads of JDBCResultSet and StatementImpl objects are in the heap. Setting the maxAge in the datasource configuration (context.xml) in tomcat helped.

    0 讨论(0)
提交回复
热议问题