I have a web site running on Arvixe's shared web hosting, the response times are atrocious, pingdom reports 10s > 15s average for the keep alive route (/keepalive)!
I'm running a precompiled 1.6 with changeset (7171b10f15b6) #19191: Fixing dynamic compilation, MachineKey is set in the web.config, trust level is set to full and the Orchard.Environment.Extensions.ExtensionMonitoringCoordinator is
disabled in the HostComponents.config.
Syscache module is on, warmup module is on, cache is on and set for 600, output cache to 900, keep alive module is also enabled.
Full file/folder permissions are set to app_data, media, modules and themes.
There are irregular but frequent delays on page loads (hence, I believe, the high average response time).
Cache does not last for more than 30s at a time, the way I know this is to open an unauthenticated session to the front-end on another browser, and to keep refreshing /Cache/Statistics until the pages in cache clear, they never last for longer
My first though was that Arvixe has app pool's recycling quickly, but through talks with support they guaranteed me that the app pool is only recycling every 60m.
With mini profiler on, every few requests there's a 4s to 5s load time reported by the profiler, but it's inconsistent, which to me it indicates a re-compile and/or app recycle.
So I'm at a loss, what could be triggering the reset of the app pool and/or invalidating syscache or cache? Is there a way to trace this on production? As anyone else experienced this slow response times from orchard on arvixe hosting?
Any help would be greatly appreciated!