Hi, Thanks for your explanation.
There are couple of alternate options apart from sending alerts to Cloudwatch and you might want to explore them if it seems valid for your case.
1. Add the Rackspace response team's email id(s) as user(s) in your Site24x7 account. By doing so, alerts can directly be sent to them instead of routing it through cloudwatch. You can even assign the alerts to be sent to separate user alert groups for different attributes. All alerts are logged in our Alert logs page.
To add a user goto Admin -> User and Alert Management -> click on the "+" icon to a user. You can create a rackspace user group and add the users to that group.
2. As a step further you can automate the actions (restart a windows service in your case) that needs to be taken when a threshold is breached. This way, you need not wait for the response team. Site24x7 will automatically execute the action for you.
a). Define the automation under Admin -> IT Automation. I have created an automation to restart a java service.
b). I will use the above action to be executed when the memory in my server crosses 65%. I have also set the poll count to be 3 so that I make sure the its a genuine problem that demands a service restart.
Product Manager, Site24x7