David Crane
08-11-2010, 09:29 AM
I should be able to get this answer from traceroute, but it is reporting "* * *" for each probe, which probably means they are timing out due to something related to our hosting environment (3tera provides hosting, so I do not have direct access to the xen layers).
Our AppLogic 2.1.1 installation is using 3 separate apps. The "www" app connects to the "data" and "search" apps, where our relational database and search engines reside. The connections from "www" go out through a NET component and arrive in "data" or "search" through IN components. The connections are done through hostnames that are DNS resolved to IP addresses in the IP block assigned to our grid.
Recently I've begun measuring the Mbps used by our production components and realized that passing "data" and "search" traffic through so many components was wasteful. Then I thought to verify whether the app-to-app hops seemed to be routed through the LAN at the border router. That's when I got stumped by traceroute.
Our AppLogic 2.1.1 installation is using 3 separate apps. The "www" app connects to the "data" and "search" apps, where our relational database and search engines reside. The connections from "www" go out through a NET component and arrive in "data" or "search" through IN components. The connections are done through hostnames that are DNS resolved to IP addresses in the IP block assigned to our grid.
Recently I've begun measuring the Mbps used by our production components and realized that passing "data" and "search" traffic through so many components was wasteful. Then I thought to verify whether the app-to-app hops seemed to be routed through the LAN at the border router. That's when I got stumped by traceroute.