Personally, I was amazed that the nodes do not affect each other. Looking at the expected lifetime of the individual stranitsy- buffers. One buffer chokes data, the other is not actually being used.
There is a hypothesis that can be disabling node architecture to get the win due to more efficient use of the buffer. But this increases the interprocessor communication.
It will be necessary to conduct an experiment, at the end I will let you know.
I know this post in a bit dated, but here I am trying to steal the SQL pertaining to sys.dm_os_waiting_tasks .
Personally, I was amazed that the nodes do not affect each other. Looking at the expected lifetime of the individual stranitsy- buffers. One buffer chokes data, the other is not actually being used.
There is a hypothesis that can be disabling node architecture to get the win due to more efficient use of the buffer. But this increases the interprocessor communication.
It will be necessary to conduct an experiment, at the end I will let you know.