By using AWS re:Post, you agree to the Terms of Use

AWS Workspaces Latency

0

Hello AWS Experts!

First and foremost, if you have stumbled upon my inquiry and are ready to aid me in resolving the issue, I appreciate you.

My company has established an AWS Workspaces environment in the Virginia region. The Workspaces service is located in two private subnets within our Virginia VPC.

When a U.S.-based user connects to the Workspaces environment, they do not deal with latency issues. Additionally, from Workspaces, the user will connect to a private EC2 instance using AWS System Manager (SSM) port forwarding. The user deals with slight latency when port forwarding/RDP into the instance. If there is a way to increase network speeds for SSM port forwarding, please inform me.

Furthermore, this is where your phenomenal expertise is most needed; our users in Bulgaria experience high latency when attempting to connect to their Workspaces and even higher when trying to use SSM to connect to their EC2 instance. When a Bulgarian user uses Workspaces, their round trip time averages around 150ms. I had them test their local network bandwidth, which most averaged around 84 to 120Mbps download speeds and 140 to 160Mbps upload speeds. Additionally, they checked their "connection status..." from the following website:

https://clients.amazonworkspaces.com/Health

The site stated that Frankfurt provided the best connection; however, we do not want to create an entirely new VPC in the Frankfurt region for Workspaces. And have our customers question why a European IP is trying to connect to their assets. A VPN may be a possibility, but we're hoping for a better solution.

I have tried the following, which non proved to improve network speeds for U.S. or Bulgarian members.

  • Established an Internal Load Balancer
  • Established VPC Endpoints
  • Established Global Accelerator

So, how could one increase network speeds to Workspaces and SSM port forwarding?

Thank you!

1 Answer
0

Round trip time (RTT) is a matter of physics and protocols must obey the law of physics - Users in Bulagria must traverse the globe no matter what. Amazon WorkSpaces is a fully managed service, that means that you are connecting to the public endpoints directly that are managed by AWS. Therefore you cannot place a load balancer in front of the managed public endpoints for the managed service. If shortening the distance vis-a-vie another region is out of the question, you can try launching a WorkSpace with the WSP protocol (instead of PCoIP) to see if that performs better for those users as WSP was conceived with higher RTT in mind.

EXPERT
answered 6 months ago
  • Hi StevieStets,

    Thank you for your response!

    I had a few Bulgarian users try the WSP protocol; however, the change was not significant. Network speeds were relatively the same. Would you recommend to create an AWS Workspaces environment in perhaps, Frankfurt?

  • The laws of physics/speed of light being what they are, you're trading latencies either way - it depends on what your users are willing to accept. It might be more acceptable that their WorkSpace is more performant while their SSM/RDP experience is less performant, compared to the WorkSpace being less performant but their SSM/RDP being more performant (sort of - it'll be hard for them to experience it since the WorkSpace itself feels sluggish. Fortunately, since it's a fully managed service, it should be pretty easy to set up WorkSpaces in Frankfurt to test with some users.

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions