> >
> > As I said earlier the method of indexing the peers based on their IP is
> > broken, not only because we do not sort the peers on their IP but also
> > because there may be multiple peers on the same IP or same IP:PORT
> > even..
> >
Yes , I knew. For the time beeing, I'm only trying to reach the point on Squid
3 for the same funcionality on Squid 2, even with the wrong IP indexing
schema
> > Regards
> > Henrik
>
> Yeah, I though Rafael was fixing that after our earlier discussions
> about keys to use. I guess not (yet) though.
>
Then we willl open the discussion on how to index the new table with several
peers on the same node...(taking a look on how an iftable does...) Just
walking on a "solid" implementation.
> The order still appears to be broken even if there was another level to
> the key. The IPA being out of order regardless of how they are configured.
> My reading of the code indicates its a linked-list of peers, right?
> If so they can/should be sorted into position using < or > operators on
> IPAddress to get around this simple case to a large degree. It will
> still be needed anyway if the key gets expanded later.
>
>
We will see.
> Good to know about that -cC trick.
>
> Amos
Received on Sun Nov 25 2007 - 10:49:08 MST
This archive was generated by hypermail pre-2.1.9 : Sat Dec 01 2007 - 12:00:05 MST