> we've also started a conversation with RackSpace netadmin types to determine whether there might be steps we could take to improve bandwidth.
RackSpace ran some tests and provided us with some traceroute/ping times between our server and 'Net locations around the world. They claim that resulting times are within their expected thresholds. However, they seem slow to us. This may be a shot in the dark, but, does anyone on this list have a resource/URL/etc illustrating average traceroute/ping times between high-end Web hosting facilities and 'Net locations around the world? Basically, I need something authoritative I can use to prove/disprove RS' claim that times are "good enough." On Wed, Oct 5, 2011 at 7:16 PM, John M Bliss <[email protected]> wrote: > True...but issue at hand is, literally, one high-traffic page that takes a > url variable, uses it to look up a row in the DB, and returns a tiny bit of > HTML. That's it. Not a whole lot of room for improvement there. > > As an aside/follow-up, further research did result in an interesting > finding: > > We use pingdom.com to test said page once per minute from servers all > around the world. Average response time (end-to-end) is approximately 260 > ms. I was tasked with reducing that number. > > We purchased and installed FusionReactor and one of the many things it > illuminated today was: average *server* time (elapsed milliseconds from when > CF first "hears" request until CF has passed HTML back to IIS) is about > 20-30 ms! > > So, while it's useful for me to screw with JVM, etc to reduce that average > (and, more importantly, to make sure that we don't have "spikes" where page > consumes lots more server time), we've also started a conversation with > RackSpace netadmin types to determine whether there might be steps we could > take to improve bandwidth. > > Thanks for all of your help! > > > On Wed, Oct 5, 2011 at 4:21 PM, Maureen Barger <[email protected]> wrote: > >> >> There's also tons of profilers out there for the Sun JVM. Rather than >> throw >> spaghetti at the wall and see what sticks, install one and watch your >> app(s) >> for a while. Perhaps your RMI interval could be longer, GC method be >> different, heap size smaller or larger etc. >> And on the other side of the coin, if you are concerned with efficiency, >> code can almost always be improved too. You can have the fastest Ferrari >> in >> the world but if the driver doesn't know how to drive stick, the ride is >> going to suck. >> Good luck! >> >> On Wed, Oct 5, 2011 at 17:02, Dave Watts <[email protected]> wrote: >> >> > >> > > Nothing in server.log available in CFAdmin. Just the successful start >> > info >> > > from after I reverted the JVM version. Am I missing something? >> > >> > The log you're interested in is one of the JRun logs file >> > (cfusion-out.log or cfusion-event.log, I forget which.). I'm not sure >> > where that's located if you're using the standalone install of CF, but >> > on multiserver, it's something like: >> > >> > c:\jrun4\logs\ >> > >> > or >> > >> > /opt/jrun4/logs >> > >> > Alternatively, you might try running CF as an application rather than >> > a service, which will pipe startup errors to your command prompt or >> > terminal window. >> > >> > Dave Watts, CTO, Fig Leaf Software >> > http://www.figleaf.com/ >> > http://training.figleaf.com/ >> > >> > Fig Leaf Software is a Veteran-Owned Small Business (VOSB) on >> > GSA Schedule, and provides the highest caliber vendor-authorized >> > instruction at our training centers, online, or onsite. >> > >> > >> >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~| Order the Adobe Coldfusion Anthology now! http://www.amazon.com/Adobe-Coldfusion-Anthology/dp/1430272155/?tag=houseoffusion Archive: http://www.houseoffusion.com/groups/cf-talk/message.cfm/messageid:348071 Subscription: http://www.houseoffusion.com/groups/cf-talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/groups/cf-talk/unsubscribe.cfm
