What is acceptable network latency in Milliseconds for users coming across a WAN?

[kappyninewood]

I've heard number like 125ms being a number to shoot for for VDI and good user experience, but what is the real world number?

[mittim12 · Accepted solution]

Acceptable latency is the number right before all of your users start to complain       This  document, http://www.vmware.com/files/pdf/view/VMware-View-5-PCoIP-Network-Optimization-Guide.pdf, says shoot for under 250 ms.   

[eeg3]

mittim12 is spot on in regards to users.

It'd be interesting to try PCoIP with 250ms latency. I have to imagine it'd be somewhat noticable, but as above, it seems they've tested it at that and the experience was solid. It'd be interesting to try the test they did.

[royberk]

Weve had users across the WAN with 80ms saying its good.

When we had some WAN issues our latency was going to from 250ms +++ and users were complaining.

It all depends on who/what you use and Im sure as the previous posters mentioned its on a business by business basis.

[Linjo]

I have customers with some users are around 200-250ms, its not pretty but for the usecase its ok.

The biggest concern is not the display-performance, its the feedback from mouse and keyboard since it will be about 500ms before the actual click or keypress gets confirmed with the users eye.

PCoIP handles latency pretty good as long as it has a lot of bandwith, high latency and low bandwith is not funny and should be avoided if possible.

There is some more optimization also coming in View 5.1 that would improve this.

// Linjo

[sketchy00]

We have some Software Developers working remotely in Argentina, connecting to our Seattle office, and often experience 280+ms latency.  Funny part is, they love it.  ...Because even with 280ms latency, PCoIP is so much better than RDP over a PPTP based VPN.

I wrote about it here:

http://itforme.wordpress.com/2012/01/05/vdi-for-me-part-1/