sibing, cache digests and NetDB
Squid's network measurement database is designed to measure the proximity of origin servers. In other words, by querying this database, Squid knows how close it is to the origin server.
Think about that
web server A& parent squid B(located in China), child squid C(in China), child squid D(in USA)
or
web server A & parent squid B(located in China), parent squid C(in China), parent squid D(in USA)
If a American user visits squid B and didn't hit, B will query the cache digests and find what the user want is both in C and D. Then B forward it to the closest squid server. I think it will always forwad to squid C (as C is more close to the origin server), but D is better to the user.
Is it right?
Can squid do that choose sibing according to the region or how close the network from user to the origin server?
Thanks,
Allen
Squid's network measurement database is designed to measure the proximity of origin servers. In other words, by querying this database, Squid knows how close it is to the origin server.
Think about that
web server A& parent squid B(located in China), child squid C(in China), child squid D(in USA)
or
web server A & parent squid B(located in China), parent squid C(in China), parent squid D(in USA)
If a American user visits squid B and didn't hit, B will query the cache digests and find what the user want is both in C and D. Then B forward it to the closest squid server. I think it will always forwad to squid C (as C is more close to the origin server), but D is better to the user.
Is it right?
Can squid do that choose sibing according to the region or how close the network from user to the origin server?
Thanks,
Allen
_______________________________________________ squid-users mailing list squid-users@xxxxxxxxxxxxxxxxxxxxx http://lists.squid-cache.org/listinfo/squid-users
