hi TS:
I'm a new user of ts, I want to build a cdn cluster for my company use.
Now I build a cluster for test use 3 machines, I want this 3 machines all as siblings,
if one miss the request, it should use icp to check whether the other 2 siblings had been
cached the quested resource.
In my records.config something abouts icp configuration as below:
CONFIG proxy.config.icp.enabled INT 2
CONFIG proxy.config.icp.icp_interface STRING eth0
CONFIG proxy.config.icp.icp_port INT 3130
CONFIG proxy.config.icp.multicast_enabled INT 1
CONFIG proxy.config.icp.query_timeout INT 1
and one icp.config as below:
localhost:192.168.1.10:3:80:3130:0.0.0.0:1
host1:192.168.1.20:2:80:3130:0.0.0.0:1
host2:192.168.1.30:2:80:3130:0.0.0.0:1
but on test, although 192.168.1.20 had beed cached the resource, 192.168.1.10 not use the
icp but back to the origin server to get the source.
I used tcpdump to detect wether there were some package on udp port 3130,but I found nothing.
In records.config there exist some note:
########################################################################
# ICP Configuration. NOTE! ICP is currently broken NOTE!
########################################################################
it says the icp is broken, but I searched this , https://issues.apache.org/jira/browse/TS-32
it seems that somebody had fixed the icp function.
I'm confued whether the icp can work or the icp doesn't work due to my wrong configuration.
Please advise and many thanks. |