You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@iotdb.apache.org by "Jialin Qiao (Jira)" <ji...@apache.org> on 2020/02/23 03:36:00 UTC

[jira] [Closed] (IOTDB-510) A NullPointException in the TsFileProcessor when flushing memtables

     [ https://issues.apache.org/jira/browse/IOTDB-510?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jialin Qiao closed IOTDB-510.
-----------------------------
    Fix Version/s: 0.10.0-SNAPSHOT
       Resolution: Fixed

> A NullPointException in the TsFileProcessor when flushing memtables
> -------------------------------------------------------------------
>
>                 Key: IOTDB-510
>                 URL: https://issues.apache.org/jira/browse/IOTDB-510
>             Project: Apache IoTDB
>          Issue Type: Bug
>          Components: Core/Engine
>    Affects Versions: 0.10.0-SNAPSHOT
>            Reporter: Xiangdong Huang
>            Assignee: Xiangdong Huang
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 0.10.0-SNAPSHOT
>
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> Sometimes when we run TravicCI, the TsFileProcessorTest will hang up. 
> If we add log in TsFileProcessor, we can find logs like:
>  
> WARN [2020-02-22 02:45:54,092] [main] org.apache.iotdb.db.engine.storagegroup.TsFileProcessor:289 - C:\Users\travis\build\jixuan1989\incubator-iotdb\server\target\data\testUnsealedTsFileProcessor.tsfile has spent 60s for waiting flushing one memtable; 1 left (first: org.apache.iotdb.db.engine.flush.NotifyFlushMemTable@29c53c4f). FlushingManager info: TSProcessors in the queue: 0, TaskPool size 0 + 0, 
> WARN [2020-02-22 02:46:54,095] [main] org.apache.iotdb.db.engine.storagegroup.TsFileProcessor:289 - C:\Users\travis\build\jixuan1989\incubator-iotdb\server\target\data\testUnsealedTsFileProcessor.tsfile has spent 120s for waiting flushing one memtable; 1 left (first: org.apache.iotdb.db.engine.flush.NotifyFlushMemTable@29c53c4f). FlushingManager info: TSProcessors in the queue: 0, TaskPool size 0 + 0, 
> WARN [2020-02-22 02:47:54,108] [main] org.apache.iotdb.db.engine.storagegroup.TsFileProcessor:289 - C:\Users\travis\build\jixuan1989\incubator-iotdb\server\target\data\testUnsealedTsFileProcessor.tsfile has spent 180s for waiting flushing one memtable; 1 left (first: org.apache.iotdb.db.engine.flush.NotifyFlushMemTable@29c53c4f). FlushingManager info: TSProcessors in the queue: 0, TaskPool size 0 + 0, 
> WARN [2020-02-22 02:48:54,122] [main] org.apache.iotdb.db.engine.storagegroup.TsFileProcessor:289 - C:\Users\travis\build\jixuan1989\incubator-iotdb\server\target\data\testUnsealedTsFileProcessor.tsfile has spent 240s for waiting flushing one memtable; 1 left (first: org.apache.iotdb.db.engine.flush.NotifyFlushMemTable@29c53c4f). FlushingManager info: TSProcessors in the queue: 0, TaskPool size 0 + 0, 
> WARN [2020-02-22 02:49:54,132] [main] org.apache.iotdb.db.engine.storagegroup.TsFileProcessor:289 - C:\Users\travis\build\jixuan1989\incubator-iotdb\server\target\data\testUnsealedTsFileProcessor.tsfile has spent 300s for waiting flushing one memtable; 1 left (first: org.apache.iotdb.db.engine.flush.NotifyFlushMemTable@29c53c4f). FlushingManager info: TSProcessors in the queue: 0, TaskPool size 0 + 0, 
> WARN [2020-02-22 02:50:54,099] [main] org.apache.iotdb.db.engine.storagegroup.TsFileProcessor:289 - C:\Users\travis\build\jixuan1989\incubator-iotdb\server\target\data\testUnsealedTsFileProcessor.tsfile has spent 360s for waiting flushing one memtable; 1 left (first: org.apache.iotdb.db.engine.flush.NotifyFlushMemTable@29c53c4f). FlushingManager info: TSProcessors in the queue: 0, TaskPool size 0 + 0,
>  
> Actually it is caused by a nullpointException but we omit the exception in the releaseFlushedMemTable() method...
>  
> The NullPointException happens when:
>  # A: there is an async flushing task;
>  # B: users call TsFileProcessor.syncClose;
> The process is:
>  # B set `shouldClose = true` and has not put the final memtable (maybe a signalMemtable) into the flushingMemtable in asyncClose() (which is called by syncClose()).
>  # A checks the `shouldeClose == true && flushingMemtables.isEmpty() ` in flushOneMemtable(), then A call endFile() to close the fileWriter; 
>  # B then put the final memtable into the flushingMemtable. When begin to flush the memtable, nullPointException is thrown (because the file Writer was set to null in endFile()).
>  
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)