Please don't reply to lustre-devel. Instead, comment in Bugzilla by using the 
following link:
https://bugzilla.lustre.org/show_bug.cgi?id=11089



(In reply to comment #4)
> Created an attachment (id=9151)
Please don't reply to lustre-devel. Instead, comment in Bugzilla by using the 
following link:
 --> (https://bugzilla.lustre.org/attachment.cgi?id=9151&action=view)
> add per-export nid and uuid proc entries
> 
> This patch adds "nid" and "uuid" entries to the per-export proc directories, 
> so
> 
> that users can at least determine what the nid is.
> 
> Nobody from ORNL seems to be subscribed to this bug - mjmac, do you know if
> they
> are still interested?  
> If we can get someone to say the additional "nid" file within the per-export
> dir
> is acceptable, I'll go ahead and backport this to b1_4.
>       

Some comments I have after going through the history of this bug entry. 

-- Agree to organize per client stat struct's in an OBD-neutral way
-- There is a need to zero brw in function, init_brw_stats
-- There is a bug in create/delete. Our fix is different. Not sure if CFS fix as
posted in Comment #4 is based on the patch posted in Comment? 
-- Fixes for the above two are posted as patch #id=9169.
 
Overall: We need to sync on our efforts. Our fixes have been tested and running
on an 80-node linux system for about 2 months. It would be great if we could
converge these into a final patch for cray machines at ORNL.

_______________________________________________
Lustre-devel mailing list
[email protected]
https://mail.clusterfs.com/mailman/listinfo/lustre-devel

Reply via email to