[openib-general] OFED 1.1 release - schedule and features

Tziporet Koren tziporet at mellanox.co.il
Thu Jul 13 02:57:47 PDT 2006


Roland Dreier wrote:
> To me this schedule seems too short to expect real new features like
>   
Since the technical details were not in this mail some features seems a 
lot of changes while they are not (to be explained).
>  > -         HCA fatal - full flow support
>   
We just want to use the current flow that exists today and test it, so I 
can declare its working (this flow was not tested thoroughly in OFED 1.0)
We also want to add a user level daemon that will catch the fatal event 
and restart the driver once it happened.
So, I guess we will find some bugs in the flow but no new features 
should be implemented.
>  > *          IPoIB
>  > -         Bonding - for high availability
>
> that have had no work done (in public at least) yet to be integrated.
> If 1.1 is going to go to code freeze in 19 days then I don't see how
> these features can be finished, let alone stabilized in time.
>   
The solution we wish to add is a not the bonding kernel module but a 
user level application that will handle failover between IB interfaces 
(sorry for not being accurate in the feature definition).
So in any case it will not harm stability of the exiting code. The 
add-on functionality may be less stable (beta) and not GA as other modules.
> In my opinion it would be better to focus on polishing what was
> already in OFED 1.0 for OFED 1.1.  Of course we should work on
> valuable features like IPoIB bonding, but I don't think they should be
> tied to OFED 1.1.
In general I agree and this is the motivation over most changes and 
features.

BTW - we will need your help regarding SRP daemon changes that were 
implemented by Ishai for OFED 1.0 but have not make it then.

Tziporet




More information about the general mailing list