IIS - remove the word 'Total' from the Metric name for Http Method Requests since filters become harder when our SiteName dimension for agregate of all sites is also called 'Total'

CloudWatch only accepts maximum of 20 data metrics per request. We therefore split any incoming metrics in to groups of 20 before sending.

Register new IisServerSiteTrafficCountLister with console and windows-service apps

New IIS traffic count reading. Will read amount of (total method) requests per second for each site in IIS. Will return null on 0, to prevent over-logging of unused sites.

For performance counters, differentiate regular value reading, and when we should divide the value by the number of processors in the local machine

Windows Service - more action logging to event log, and do not log every single status report, as this means we're storing all the stats in the event log!

ignore .config files. If you want to commit some sample config, add the file manually

I'm afraid we can't call any Console methods inside the manager as this causes incompatibilities where there is no Console (ie: when in the scope of a windows service)

Windows Service runs manager on a new thread on start

IIS provider: initial thought on providing traffic counts

Windows Service (basic dependency injection so far). Note that amazon credentials are fake and will not work - replace with your own.

CounterManager - Check for null on log writer action when writing to log

Initial support for specifying if Update of counter lists are required by a counter list provider.

When dealing with Windows PerformanceCounters, read the value upon intialisation of the reference so that counter is initialised and the first value sent to CloudWatch is not 0

Change iteration interval default to 60 seconds.