We are currently working on a new set of diagnostics tools that will be available as part of the Diagnostics module. Here's a short summary of the new capabilities
Page Compilations Profiler - allows you to find pages that compile frequently and therefore take a lot of time to render to the user
HTTP Requests Profiler - allows you to identify the requests to your server that take too long and where the time goes - compilation, execution, database communication
Background Tasks Profiler - allows you to identify background tasks and the time they take to execute
N+1 Queries Profiler - allows you to identify pages that execute a number of similar (and redundant) queries to the database
All these will come together with the existing SQL Queries Profiler and will be possible to enable/disable independently, so that you can keep only the ones you want.
In addition, we will also have an Insights dashboard with valuable insights like number of pages that are compiled frequently or HTTP requests that take too long.
You can see a prototype of these features at the following link:
Please shoot your questions as well as any ideas for insights and performance tips you'd like to see.
The only time I tried to use the diagnostics it didn't work because my DB had a period in the name... has that been fixed yet?
(Not to derail at all, these are definate improvements...)
Will the HttpRequests be hooking into ServiceStack as well, or just the oData stuff...
It should work with a period too and we will double-check that for you :)
The HTTP requests will catch all requests that go into Sitefinity - pages, scripts, api calls, etc.
It will be great if you create profiler for cold startup. Because now, startup process looks like a black box. For example:
-page compilation 50sec
-load time of each module
News 10 sec
Sitemap 2 sec
DynamicModule1 3 sec
Dynamic module2 5 sec
I personally haven't used the existing diagnostics module. Yet from the description of these new items, I probably definitely would to help test slowdowns of things.
And I do agree with Steve, that a backend status/management page of background tasks would be super beneficial.