Uptime.com Blog
Website and Performance Monitoring for Edge Cases
Specific needs are compelling but also hard to plan for. Your use case may be the reason you are searching for a monitoring provider, but the ability of your provider to adapt to your edge cases will be the reason you stay.
The challenge is in discerning if a provider will be able to rise to meet your needs in unknown circumstances.
In monitoring, there are some uniform needs. Everyone wants to know if their site is UP so HTTP(S) checks meet use cases universally. For most sites, shopping carts and login forms need monitoring too, so synthetic monitoring tools like Transaction Checks are also widely used.
Status Pages are becoming more popular (and more necessary), and so are scheduled reports in response to increasing SLA accountability. But what about industry specific performance monitoring needs, server networks, and support availability?
How close are you to the edge?
Testing Third-Party Providers
Integrating with third-party services like payment providers, hosted contact forms, and POS systems is the least edgy edge case. Creating all of your needed operations and managing them in-house is not always practical – but those important pathways can be expensive to monitor.
Transaction Checks are great for simulating user actions but if your provider charges you per transaction or form submission (even if they’re test users) checks running even at 60-minute intervals 24/7 can be costly.
A solution here can be to set up a one-two punch check combo with an HTTP(S) check and an API check to first monitor that the URL is up (HTTP), and then create an API check to post and validate data on the provider side. Status endpoints work well for this type of operation, because you can usually freely query to make sure the provider is up.
You could use an API check to first validate for 200 OK, but the benefit of two checks in this edge case is more specific alerting. If the HTTP(S) fails – your site is down. If the API check fails – your third party service may be having issues.
Monitoring Performance from Specific Country Locations
The ability to configure probe server locations from which your monitoring checks run can be a deciding factor in choosing a monitoring provider – not just in terms of being able to monitor where your traffic is coming from, but also in meeting any mandated requirements.
Government organizations especially need to be able to provide direct monitoring coverage for their specific countries and regions – and may need to exclude other locations from their monitoring infrastructure.
Load Balancers & Monitoring Hosts
Load balancers can create some sticky sessions (where an end user’s requests are always directed to the same server). This dominant server serves as an entry point, but it’s also important to monitor the hosts sitting behind it.Using a simple HTTP check, you can monitor the load balance IP address with added host headers that provide visibility into each of the back end servers. The end result is a notification when any of your load balance servers fail.
In monitoring terms, configured probe servers may always be directed to the same “entry” server – missing issues that may be affecting other servers behind it.
Curious how this is done? Check out the Uptime.com way.
Monitoring your CDN
A CDN (content delivery/distribution network) is a network of proxy servers and their data centers that globally distributes data from a single origin server by caching data near (geographically) where end users are accessing the internet.
CDN monitoring can be configured with one of Uptime.com’s basic and most widely used check types, HTTP(S). Recently upgraded, V2 supports check creation for important CSS or Javascript assets being served from the CDN before running the check from multiple locations triggering different CDN edge servers.
Want to try it out? Create an Uptime.com HTTP(S) check and utilize String to Expect to look for important assets that the CDN is responsible for loading.
With CDN’s global nature, this sequential multi-location testing is a reliable way to determine if your CDN is UP.
Monitoring Edge Cases: The Bottom Line
Uptime and performance monitoring is an expanding universe and monitoring can be tricky. The more elements added to and integrated with your site – the more chances there are for something to go down. From your site’s foundation and third-party providers to your dynamic marketing elements and location requirements, there are an infinite number of moving parts making adaptation and 360º monitoring key traits to seek out in your monitoring solution.
The other trait is communication. We’ve got some of the best support around for when you get stuck. Test us! Reach out and see if we’re the right fit for your edge case.
Minute-by-minute Uptime checks.
Start your 14-day free trial with no credit card required at Uptime.com.
