The kernel tries to divide interrupt traffic evenly
across the processors. But, on what basis
kernel tries to divide interrupt traffic evenly
across the processors ? How is the load sharing
of interrupts done ? Where can i get the related info ?
Thx in advans,
Karthik Balaguru
> The kernel tries to divide interrupt traffic evenly
> across the processors. But, on what basis
> kernel tries to divide interrupt traffic evenly
> across the processors ? How is the load sharing
> of interrupts done ? Where can i get the related info ?
For the Linux user-space IRQ balancer, look here:
http://irqbalance.org/
DS
But the userspace IRQ balancer seems to be a daemon
that is used only incase the IRQBALANCE is disabled.
Strange !
Its strange that IRQBALANCE should be disabled to
make the interrupts to be distributed/shared between the various
processors. But, if the IRQBALANCE is enabled, all the interrupts are
being processed by one processor only.
I think, the term IRQBALANCE should be changed to
IRQ_UNBALANCE. Any thoughts ?
From the link, http://irqbalance.org/documentation.php , i got the
below lines that clarify regarding the need for the new irqbalance,
But, why does it remain separate and why is still the obsolete
irq balancer present in the kernel code ? Any ideas ?
" The current Linux irqbalance program is several years old
in design, and is blissfully unaware of the ideas of Quad
(or even Dual) core or even power usage. The program is
conceptually closer to the naive balancing than to the smart
interrupt balancer. "