Some NFS file transfers fail and hang automounting

From Linux NFS

(Difference between revisions)
Jump to: navigation, search
(Created page with '== Symptoms == * Given a file "foo" of 50Mb on a remote machine "remote". * This command will never finish <pre><localmachine $> cp /remote_mount_point/foo bar </pre> * bar will…')
Line 1: Line 1:
== Symptoms ==
== Symptoms ==
* Given a file "foo" of 50Mb on a remote machine "remote".  
* Given a file "foo" of 50Mb on a remote machine "remote".  
-
* This command will never finish <pre><localmachine $> cp /remote_mount_point/foo bar </pre>
+
** This command will never finish <pre><localmachine $> cp /remote_mount_point/foo bar </pre>
-
* bar will have a size less than foo.
+
** bar will have a size less than foo.
-
* automounting of the local machine is hung.
+
** automounting of the local machine is hung.
-
The error message in /var/log/messages is:
+
* The following message will show up in /var/log/messages
-
kernel: Callback slot table overflowed
+
<pre>kernel: Callback slot table overflowed</pre>
If foo is lesser than 10Mb, the problem does'nt occur.
If foo is lesser than 10Mb, the problem does'nt occur.

Revision as of 14:42, 22 October 2010

Symptoms

  • Given a file "foo" of 50Mb on a remote machine "remote".
    • This command will never finish
      <localmachine $> cp /remote_mount_point/foo bar 
    • bar will have a size less than foo.
    • automounting of the local machine is hung.
  • The following message will show up in /var/log/messages
kernel: Callback slot table overflowed

If foo is lesser than 10Mb, the problem does'nt occur. After several tries with the source of 50Mb, the different sizes of the target file "bar" are random.

This incident occurs with: sunrpc.tcp_slot_table_entries = 16

but after tunning this parameter, all is fine with: sunrpc.tcp_slot_table_entries = 32

Personal tools