I am presently attempting to build (a less complicated version of) a standing website much like Amazon's Service Health Dashboard
Essentially, I have to connect within 3 Data Centers and a few websites with a minimum of an Up/Lower message.
I am mostly acquainted with C# plus some web programming. I am slightly acquainted with web services, also it appears like that might be a great way to implement this. I'm able to Google all things in lessons for individual parts, but I am unsure the way the summary of something similar to this works.
Generate an income comprehend it is it might be divided in to these parts:
- Status Website - ASP.Internet
- Web Service - C#/ASP.Internet
- Some type of program to transmit status to the net service - C#?
For that website status, the status site only will operate a script to see if a URL returns a mistake.
I am confused around the simplest method to implement the information Center status. From things i understand, I'm able to write an internet service which get reactions in the Data Center. May be the simplest method of doing mtss is a C# program situated in every Data Center that inspections the status (most likely logging right into a SQL database) after which delivering the status to the net service? I am also confused around the simplest method of getting these details towards the status website. I suppose the simplest way could be for that web plan to email personal files that the status website would read. Also, I suppose it will have to time the net service calls and benefit by an outage whether it does not get a status message in some time.
Any high-level overview help could be greatly appreciated. Thanks.
The simplest way is always to begin with your meaning of health.
Decide what criteria, for every element of the information center, define health. Figure out how to measure individuals criteria, ideally from the reliable server within the data centers. Observe that It is best to learn to measure them outdoors of the web service, since you are unfamiliar with web services.
Produce a web plan to operate on that reliable server to measure individuals criteria. Then attach the measures for your web service, after determining the way you should present the outcomes back to the net site.
The site could then make use of the is a result of calling these types of services to update a standing dashboard. It might use AJAX to call the help, on all of the data centers, asynchronously.
This will get a little more complicated if you want to incorporate actions in line with the status from the data center. For example, you may need a button to begin a database when the database is discovered to be lower.
EDIT: I should also make certain that you're conscious of items such as the System Center items from Microsoft. They might be overkill for which you are doing, however they do give a standard, flexible mechanism for monitoring data center status, as well as for following through to fix problems. Also, see http://www.codeplex.com/dfo.
I additionally desired to point out that IPSwitch constitutes a product known as "WhatsUpGold" that does exactly this. I have used it within the last four years also it can most likely do all you need as they are. Most significantly, it is extremely economical! I am a version behind, since i have did not need the additional features! It supports a distributed multi-site network (agent each and every site monitors status/health insurance and reviews to your primary location/hub).
*As John pointed out above, the important thing to monitoring a device's status is determining it's health. Just obtaining a status of up or lower isn't enough, because so many other activities ping latency, disk utilization, cpu utilization, etc... depends in your unique atmosphere.