[lustre-discuss] File locking errors.

Michael Di Domenico mdidomenico4 at gmail.com
Tue Feb 20 09:29:16 PST 2018


On Tue, Feb 20, 2018 at 8:58 AM, Patrick Farrell <paf at cray.com> wrote:
> There is almost NO overhead to this locking unless you’re using it to keep threads away from each other’s on multiple nodes, in which case the time is spent doing waiting your app is asking for.

agreed.  my reply was slightly worded incorrectly from a contextual
point of view.  i did not intend to imply that the overheard was a
lustre issue, although from a user perspective it could be viewed as
such.

my intent of the statement was that if you turn on global locking you
will see an impact, however larger or small that impact is certainly
debatable.

the lustre wiki even mentions it: "flock: enable support for
cluster-wide, coherent file locks. Must be applied to the mount
commands for all clients that will be accessing common data requiring
lock functionality. Cluster-wide locking will have a detrimental
impact on file system performance, and should only be enabled when
absolutely required"

so i'm not off-base, but there's room for discussion.


More information about the lustre-discuss mailing list