[ofa-general] Re: [Query] ib add path record cache

[Devesh Sharma]

On 5/14/07, Sean Hefty <sean.hefty at intel.com> wrote:
> >This can be treated as a facility similar to what we have in ARP table
> >for TCP/IP. Secondly this will help in debugging of some new up-coming
> >partially infiniband complaint hardware.
>
> But unless such a path actually exists to the remote node, I don't see that it's
> useful.  And if such a path exists, I would expect it to be returned by the SA.
But initially this will generate a packet for each path, while sys
admin knows that path is there and he can hard-code the entries for
it. Other thing is that why Admin will care about creating such record
while SA is itself taking care, right?
> Can you clarify its use more wrt the subnet in general?
Again the same, in most cases Administrator knows that some path is
there between Node A and Node B, then why to introduce more delay in
making stack up by introducing extra packets (generated by
sa_cache_module). In later stages if something is changing, may be, it
will generated only few packets to update the cache.

Another point I want to know is,
When local_sa_cache module will be inserted? After SM comes up or
Before SM comes up?
I think its after SM is up, So this is introducing extra efforts for
Admin, He will have to wait for SM to come up and then insert sa_cache
module.

If Its inserted before SM is coming up (I am assuming SM is running on
some node not on switch) then First Forced schedule_update() is
waisted, and for the first application presence of cache is
meaningless. Why not to keep cache effective right from the start?
CMIIW
>
> >yes, I want them to remain in the DB, my idea is similar to the hard
> >coding of ARP table entries in TCP/IP.
> >How do you see this can be achieved?
>
> A simple flag or setting the update counter on the added path to the maximum
> should be sufficient.
>
> - Sean
>