Cassandra-节点可以'；由于java.lang.OutOfMemoryError:直接缓冲内存，无法重新启动

我正在使用最新版本接收20 GB的数据。 我有一个5节点的C*集群。接收数据的客户端应用程序正在本地节点上运行，该节点不是C*群集的一部分（而是同一LAN的一部分）。我还在我摄取到的表上使用索引。 [cqlsh 5.0.1 | Cassandra 3.0.8 | CQL规范3.4.0 |本机协议v4]

我的java客户端应用程序的工作方式如下：

json文件被解析

将生成一个PreparedStatement

我的文件中每个元素的一个BoundStatement将通过

executeAsync

发送到C*集群[注意：发送了很多行]

大约过了一半，我的两个节点死了，

/var/log/cassandra/system.log

显示

    ERROR [CompactionExecutor:1] JVMStabilityInspector.java:140 - JVM state determined to be unstable.  Exiting forcefully due to:
java.lang.OutOfMemoryError: Direct buffer memory
        at java.nio.Bits.reserveMemory(Bits.java:693) ~[na:1.8.0_91]
        at java.nio.DirectByteBuffer.<init>(DirectByteBuffer.java:123) ~[na:1.8.0_91]
        at java.nio.ByteBuffer.allocateDirect(ByteBuffer.java:311) ~[na:1.8.0_91]
        at org.apache.cassandra.utils.memory.BufferPool.allocate(BufferPool.java:108) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.memory.BufferPool.access$1000(BufferPool.java:45) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.memory.BufferPool$LocalPool.allocate(BufferPool.java:387) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.memory.BufferPool$LocalPool.access$000(BufferPool.java:314) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.memory.BufferPool.takeFromPool(BufferPool.java:120) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.memory.BufferPool.get(BufferPool.java:92) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.util.RandomAccessReader.allocateBuffer(RandomAccessReader.java:87) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.compress.CompressedRandomAccessReader.access$100(CompressedRandomAccessReader.java:38) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.compress.CompressedRandomAccessReader$Builder.createBuffer(CompressedRandomAccessReader.java:275) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.util.RandomAccessReader.<init>(RandomAccessReader.java:74) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.compress.CompressedRandomAccessReader.<init>(CompressedRandomAccessReader.java:59) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.compress.CompressedRandomAccessReader$Builder.build(CompressedRandomAccessReader.java:283) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.util.CompressedSegmentedFile.createReader(CompressedSegmentedFile.java:145) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.util.SegmentedFile.createReader(SegmentedFile.java:133) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.sstable.format.SSTableReader.getFileDataInput(SSTableReader.java:1711) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.columniterator.AbstractSSTableIterator.<init>(AbstractSSTableIterator.java:93) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.columniterator.SSTableIterator.<init>(SSTableIterator.java:46) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.columniterator.SSTableIterator.<init>(SSTableIterator.java:36) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.sstable.format.big.BigTableReader.iterator(BigTableReader.java:62) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.SinglePartitionReadCommand.queryMemtableAndDiskInternal(SinglePartitionReadCommand.java:580) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.SinglePartitionReadCommand.queryMemtableAndDisk(SinglePartitionReadCommand.java:492) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at com.stratio.cassandra.lucene.IndexService.read(IndexService.java:618) ~[cassandra-lucene-index-plugin-3.0.8.0.jar:na]
        at com.stratio.cassandra.lucene.IndexWriterWide.finish(IndexWriterWide.java:89) ~[cassandra-lucene-index-plugin-3.0.8.0.jar:na]
        at org.apache.cassandra.index.SecondaryIndexManager$IndexGCTransaction.commit(SecondaryIndexManager.java:958) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.compaction.CompactionIterator$1$1.onMergedRows(CompactionIterator.java:197) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.rows.UnfilteredRowIterators$UnfilteredRowMergeIterator$MergeReducer.getReduced(UnfilteredRowIterators.java:484) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.rows.UnfilteredRowIterators$UnfilteredRowMergeIterator$MergeReducer.getReduced(UnfilteredRowIterators.java:446) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.MergeIterator$ManyToOne.consume(MergeIterator.java:220) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.MergeIterator$ManyToOne.computeNext(MergeIterator.java:159) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.AbstractIterator.hasNext(AbstractIterator.java:47) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.rows.UnfilteredRowIterators$UnfilteredRowMergeIterator.computeNext(UnfilteredRowIterators.java:428) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.rows.UnfilteredRowIterators$UnfilteredRowMergeIterator.computeNext(UnfilteredRowIterators.java:288) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.AbstractIterator.hasNext(AbstractIterator.java:47) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.transform.BaseRows.hasNext(BaseRows.java:128) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.ColumnIndex$Builder.build(ColumnIndex.java:111) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.ColumnIndex.writeAndBuildIndex(ColumnIndex.java:52) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.sstable.format.big.BigTableWriter.append(BigTableWriter.java:149) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.io.sstable.SSTableRewriter.append(SSTableRewriter.java:125) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.compaction.writers.DefaultCompactionWriter.realAppend(DefaultCompactionWriter.java:57) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.compaction.writers.CompactionAwareWriter.append(CompactionAwareWriter.java:109) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.compaction.CompactionTask.runMayThrow(CompactionTask.java:182) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.utils.WrappedRunnable.run(WrappedRunnable.java:28) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.compaction.CompactionTask.executeInternal(CompactionTask.java:78) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.compaction.AbstractCompactionTask.execute(AbstractCompactionTask.java:60) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at org.apache.cassandra.db.compaction.CompactionManager$BackgroundCompactionCandidate.run(CompactionManager.java:263) ~[apache-cassandra-3.0.8.jar:3.0.8]
        at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511) ~[na:1.8.0_91]
        at java.util.concurrent.FutureTask.run(FutureTask.java:266) ~[na:1.8.0_91]
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) ~[na:1.8.0_91]
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) [na:1.8.0_91]
        at java.lang.Thread.run(Thread.java:745) [na:1.8.0_91]

我怎样才能让他们重新开始？ 我觉得他们试图重新构建一些排队的插入，因为我只是从我的客户那里用

writeAsync

来“垃圾邮件”插入，而cassandra最终被插入的数量压倒了

如何防止下次摄取数据时发生这种情况

---

这是由于堆外缓冲区占用了太多内存，因为旧版本的Java（1.8.102和1.9.104之前）对此没有限制：

尝试升级到较新的Java并添加：

-Djdk.nio.maxCachedBufferSize=1048576

---

要

cassandra env.sh

或

jvm.options

它仍然没有索引吗？@phact我考虑删除索引并重新运行数据摄取。但是，由于此错误，我当前无法启动5个节点中的2个。如何使它们能够重新启动？我的文件夹/var/lib/cassandra/data/*mykeyspace*/只包含名为“mytable-SOME\u ID\u HERE”的文件夹。在能够重新启动之前，还有什么需要删除/清理的吗？它在commitlog replay上爆炸了吗？您可能想暂时删除这些数据，看看它是否能让您恢复。@phact我已经删除了

/var/lib/cassandra/d的所有子文件夹‌​失败节点上的ata/*mykeyspace*/

。它们现在可以重新启动。我删除了索引并重新创建了整个表。现在将再次启动摄取并报告。@phact删除索引后，导入程序已完成，cassandra未出现任何严重错误。（除了一个节点的硬盘在午夜死亡，但我不认为这与lol有关）。我想我应该停止使用executeAsync而改用execute？system.log中充满了“已删除变异消息…”日志，因为硬盘驱动器无法跟上接收速度。这只发生在硬盘驱动器随后死亡的节点上。

-Djdk.nio.maxCachedBufferSize=1048576