Hello. I'm trying to setup traffic server cluster (as forwarding proxy) with two servers. Configurations for the cluster may be OK. Objects are successfully cached and served.
However, I found some strange statistics. On one of the servers (A), the value of "proxy.process.cache.volume_0.read.failure" has been increasing periodically without any incoming requests. Also, there are no logs on common, extend, extend2, squid, and diags. But the other server's value is not increased. It is only increased by cache miss for requests. I tried to capture packets on 8080 which the traffic server A has bound, but no packets came. The traffic server version is 3.2.4 and my OS is CentOS 6.3 x86_64. How do I find the reason? Does anyone help me? Regards, Hyangtack.
