记录一次线上OOM
发现服务宕机两次,于是查看日志,发现错误如下:
原因:
datab 15:46:59.302 [SimpleAsyncTaskExecutor-7] ERROR o.s.a.i.SimpleAsyncUncaughtExceptionHandler - Unexpected error occurred invoking async method 'public void com.chinadatab.job.ScheduledTasks.lucene_enterprise()'. java.lang.StackOverflowError: null at sun.nio.fs.WindowsNativeDispatcher.CreateFile0(Native Method) at sun.nio.fs.WindowsNativeDispatcher.CreateFile(Unknown Source) at sun.nio.fs.WindowsChannelFactory.open(Unknown Source) at sun.nio.fs.WindowsChannelFactory.newFileChannel(Unknown Source) at sun.nio.fs.WindowsFileSystemProvider.newByteChannel(Unknown Source) at java.nio.file.Files.newByteChannel(Unknown Source) at java.nio.file.Files.createFile(Unknown Source) at org.apache.lucene.store.NativeFSLockFactory.obtainFSLock(NativeFSLockFactory.java:98) at org.apache.lucene.store.FSLockFactory.obtainLock(FSLockFactory.java:41) at org.apache.lucene.store.BaseDirectory.obtainLock(BaseDirectory.java:45) at org.apache.lucene.index.IndexWriter.<init>(IndexWriter.java:727) at com.chinadatab.lucene.LuceneManage._IndexWriter(LuceneManage.java:934) at com.chinadatab.lucene.LuceneManage._IndexWriter(LuceneManage.java:940)
从日志中可以看出是那个方法导致的OOM,这里是栈溢出,于是找到对应的语句查看,有一个定时器,每隔30分钟执行一次,每次查询取了10000条数据,处理完10000完数据后批量通过indexWrite生成索引导致的,如下:
@Async @Scheduled(initialDelay=1000*60*30, fixedDelay=1000*60*30) /** * 完善基础企业库 */ public void lucene_enterprise(){ String tableName="company"; LuceneManage luceneManage= LuceneManage.getInstance(); Page page = luceneManage.get(0, 1, true, tableName); List<Map<String, Object>> lists = page.getList(); long id=0; for (Map<String, Object> map : lists) { id=Long.parseLong(map.get("id").toString()); } if(id>0) { TngouDBHelp TngouDBHelp = new TngouDBHelp(); List<Enterprise> list; try { //这里取了10000条数据,导致了OOM list = mapper.serach(id); if(list==null||list.isEmpty()) { return; } List<Fields> ls= new ArrayList<>();
//创建了大量对象 for (Enterprise e : list) { Fields fields = new Fields(); fields.add(new Field("id", e.getId()+"", Type.Key)); fields.add(new Field("person", e.getOper_name(), Type.Text)); fields.add(new Field("name", e.getName(), Type.Text)); fields.add(new Field("address", e.getAddress(), Type.Text)); ls.add(fields); }; TngouDBHelp.insert(tableName,ls); } catch (Exception e1) { e1.printStackTrace(); } } }
分析:
栈内存为线程私有的空间,每个线程都会创建私有的栈内存。栈空间内存设置过大,创建线程数量较多时会出现栈内存溢出StackOverflowError。
同时,栈内存也决定方法调用的深度,栈内存过小则会导致方法调用的深度较小,如递归调用的次数较少。
-Xss:如-Xss128k
解决:
调整JVM参数增加栈的大小,统计减少每次处理的数据量,修改为每次5000条数据