On Fri, 5 Mar 1999, pascal tscharner wrote:
> we've got some children and a couple of parents.
> the cache-digest has been enabled on the child-squids.
> on each child we've defined 2 parents with different weigths.
> if one parent is killed the child recognizes its death but he is still
> trying to access the killed parent cause of the cache-digest entries.
This should not be happening. Unless there is a bug, digests are not used
for invalid parents. How do you know that the child recognizes parent's
death? What makes you think that requests to a dead parent are due to
digests?
Thanks,
Alex.
