yang yang created HDFS-17757:
--------------------------------
Summary: An error is reported when append to file
Key: HDFS-17757
URL: https://issues.apache.org/jira/browse/HDFS-17757
Project: Hadoop HDFS
Issue Type: Bug
Components: dfs
Affects Versions: 3.4.1, 3.3.1
Reporter: yang yang
we set the following dfs parameters to improve performance:
dfs.namenode.file.close.num-committed-allowed=1
dfs.blockreport.incremental.intervalMsec=1000
We upload a dfs file /tmp/a.txt, and then immediately call the append method to
append the file，we found following error in namenode:
2025-03-04 13:15:29,974 | WARN | IPC Server handler 59 on default port xxx |
DIR* NameSystem.append:
org.apache.hadoop.hdfs.server.namenode.NotReplicatedYetException: append:
lastBlock=blk_108xxxx240_100xxx04 of src=/tmp/a.txt is COMMITTED but not yet
COMPLETE. | FSDirAppendOp.java:149
Our hadoop replicated number is set 2, we suspect that only one block is
reported successfully when appending. Is this a bug? Do we have any good ways
to solve this problem?
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-dev-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-dev-h...@hadoop.apache.org
