[GE users] running out of inodes, what triggers?

Rayson Ho rayrayson at gmail.com
Thu Oct 16 17:12:18 BST 2008


    [ The following text is in the "ISO-8859-1" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some special characters may be displayed incorrectly. ]

On 10/16/08, Eric Wu <ewu at bbn.com> wrote:
> "in high-throughput clusters classic spooling could run out of inodes".
>
> Could anybody elaborate on that?  Is running out of inodes triggered by too
> many hosts?
> Too many pending jobs?  Too many running jobs?  If so, what would
> approximate limits be?

Usually too many submitted (pending, running) jobs. The number of
inodes is a filesystem limit, so it depends on your OS/filesystem. For
some older filesystems, you need to decide how many inodes you need at
filesystem creation time.

(Google for "inode" further details...)

With Berkeley DB, the number of files used by SGE to store job
information is usually small, as each additional job only occupies
some storage space in the database file and not an actual file visable
in the filesystem.

Rayson



>
> Thanks in advance for any help.
>
>
> Eric Wu
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> users-unsubscribe at gridengine.sunsource.net
> For additional commands, e-mail:
> users-help at gridengine.sunsource.net
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list