On 11/09/2013 4:32 p.m., Alex Rousskov wrote:
Hi Amos,What is the status of your pinning_hier_note.patch referenced below? http://master.squid-cache.org/~amosjeffries/patches/pinning_hier_note.patch I know the patch helps in some cases, but I seem to recall somebody (possibly you) indicating that it needs more work. Should the fix be committed (to trunk, at least) in its current state, or is that a bad idea for some reason?
It is not in trunk already, so probably just needed testing feedback to confirm it worked properly. Its from a time before I had fully worked out how pinning was working.
With my current knowledge I think its necessary to add. Although ideally we should use the value from when the connection was pinned (in ConnStateData::pinning.host).
I will apply that shortly. Amos
