Forum Discussion
How Scalable is Windows Admin Center
We are currently evaluating WAC and I was wondering if there are any recommendations\guidelines for the number of nodes you can add?
This article suggest its not suitable for anything more than 500 - https://blog.5nine.com/windows-admin-center-strengths-weaknesses-enterprise-readiness-and-everything-else-you-need-to-know
Not sure of its suitability to manage our 20,000+ servers
Any guidance is appreciated
Thanks
Stephen
- The number of nodes in your connection list is irrelevant, since WAC only connects to one node at a time. The question of scale really only applies to the number of concurrent users using WAC when it is deployed as a gateway service. In our test automation, we typically have more than 50 simultaneous test cases running in parallel, which is equivalent to more than 50 concurrent users all performing management tasks through a single Windows Admin Center gateway at the same time. Our standard test gateway configuration is a VM with 2 virtual procs, 4 GB of RAM, 100 GB storage (non-SSD). With this configuration, we see an acceptable level of performance and response time through the gateway with no timeouts.
- Jeff Woolslayer
Microsoft
The number of nodes in your connection list is irrelevant, since WAC only connects to one node at a time. The question of scale really only applies to the number of concurrent users using WAC when it is deployed as a gateway service. In our test automation, we typically have more than 50 simultaneous test cases running in parallel, which is equivalent to more than 50 concurrent users all performing management tasks through a single Windows Admin Center gateway at the same time. Our standard test gateway configuration is a VM with 2 virtual procs, 4 GB of RAM, 100 GB storage (non-SSD). With this configuration, we see an acceptable level of performance and response time through the gateway with no timeouts.Awesome - thanks jeff
How did you achieve this? I've deployed WAC on larger servers and have poor performance connecting to virtual machines and failover clusters. The portal has port latency and take forever to full load a page. Is that the same experience you have and you just deal with it?
