Re: Adjustment of spinlock sleep delays

From: "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>

> To forestall this scenario, I'm thinking of introducing backoff into the
> sleep intervals --- that is, after first failure to get the spinlock,
> sleep 10 msec; after the second, sleep 20 msec, then 40, etc, with a
> maximum sleep time of maybe a second. The number of iterations would be
> reduced so that we still time out after a minute's total delay.

What about use the same algorithm used in ethernet when a collision is
detected?

When a collision occurs:

1) Stop sending.
2) Wait for some random amount of time between 0-T seconds
3) Try again.
4) If collision occurs again, increase T using an algorithm, go to
Step 2.

algorithm:

T = rand ( 2n -1 ) * 51.2 ms

where n- number of collisions

Regards
Gaetano Mendola