Folks,

Can anyone please give an idea around write behind store and how it
operates? Specifically how the buffers are distributed among nodes and how
copies are maintained (if any).

I am trying to find a way if we can weasel around with maintaining copies
of primary node queue as backup queues on secondary nodes (as Dmitriy
suggested).

-- 
Regards,

Atri
*l'apprenant*

Reply via email to