WangZhichao created HDFS-15056:
----------------------------------
Summary: NFS: Error 'Stale file handle' caused by executing
'mount' command in the mount directory after mounting to nfs-hdfs-gateway
Key: HDFS-15056
URL: https://issues.apache.org/jira/browse/HDFS-15056
Project: Hadoop HDFS
Issue Type: Bug
Components: nfs
Affects Versions: 3.2.1
Reporter: WangZhichao
10.43.33.246 is nfs-hdfs-gateway,
centos87 is nfs-hdfs-client,
The reproduction steps are as follows:
[root@centos87 ~]# mount -t nfs -o vers=3,proto=tcp,nolock 10.43.33.246:/
/var/data/share/am
[root@centos87 ~]# cd /var/data/share/am
[root@centos87 am]# ls
csvFile hbase oneminer solr sparkSQL spark-tmp testaie3 user zaip_data_87
hadoop hive saveModelPath spark sparkSQL-tmp test time_series_demo wzc
[root@centos87 am]# mount | grep 10.43
10.43.33.246:/ on /var/data/share/am type nfs
(rw,relatime,vers=3,rsize=1048576,wsize=1048576,namlen=255,hard,nolock,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.43.33.246,mountvers=3,mountport=4242,mountproto=tcp,local_lock=all,addr=10.43.33.246)
[root@centos87 am]# ls -lrt
ls: cannot open directory .: Stale file handle
Key point:
1. The operating system of the node(centos87) where nfs-client is located is
CentOS Linux release 7.7.1908 (Core). It can be reproduced with Redhat7, but it
can not be reproduced with Redhat6.
2. After the command 'mount' is executed in the mount
directory('/var/data/share/am'), the problem will recur. If it is executed in
other directories, the problem will not recur.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-dev-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-dev-h...@hadoop.apache.org
