im just wondering what a guideline would be, for how much usage i can handle on 1 app instance at default settings, hosted on my own windows server OS that i hire in a remote datacenter.
secondly, can i upgrade the website, or change settings to stick with 1 instance?
and what will cause the bottleneck first before i need multiple instances? and what is its name?
Your question is like asking, "How long is the string?".
Application performance comes down how well the application is designed. I feel the biggest benefit is making efficient use of the asynchronous API calls. Of course the code must be well written.
Local state is usually the biggest hurtle when moving from a single application to a farm. If the application uses local memory then moving to multiple application instances is a problem. The memory/state must be shared with all application instances.
Distributed cache solves this problem.
Member
5 Points
165 Posts
when would one have to worry about multiple instances of web applications (server farming?)
Jul 17, 2020 01:03 PM|fazioliamboina|LINK
im just wondering what a guideline would be, for how much usage i can handle on 1 app instance at default settings, hosted on my own windows server OS that i hire in a remote datacenter.
secondly, can i upgrade the website, or change settings to stick with 1 instance?
and what will cause the bottleneck first before i need multiple instances? and what is its name?
All-Star
53041 Points
23612 Posts
Re: when would one have to worry about multiple instances of web applications (server farming?)
Jul 17, 2020 02:46 PM|mgebhard|LINK
Your question is like asking, "How long is the string?".
Application performance comes down how well the application is designed. I feel the biggest benefit is making efficient use of the asynchronous API calls. Of course the code must be well written.
Local state is usually the biggest hurtle when moving from a single application to a farm. If the application uses local memory then moving to multiple application instances is a problem. The memory/state must be shared with all application instances. Distributed cache solves this problem.