Solved

Acceptable Latency for Terminal Services Session

Posted on 2006-06-26
16
4,662 Views
Last Modified: 2014-11-01
Trying to get an idea of the latency between our servers and a remote site, we had some of the users ping the server during peak internet hours.  This is the log:

Reply from XXX.XXX.XXX.XX: bytes=32 time=289ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=289ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=288ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=300ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=287ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=288ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=292ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=293ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=308ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=286ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=285ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=286ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=286ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=289ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=287ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=285ms TTL=115
Reply from XXX.XXX.XXX.XX: bytes=32 time=288ms TTL=115

Ping statistics for XXX.XXX.XXX.XX:
    Packets: Sent = 266111, Received = 264128, Lost = 1983 (0% loss),
Approximate round trip times in milli-seconds:
    Minimum = 268ms, Maximum = 548ms, Average = 285ms

Is this average RTT too much?  Is there a better way of determining that?
0
Comment
Question by:hrasheed
  • 4
  • 2
  • 2
  • +5
16 Comments
 
LVL 8

Accepted Solution

by:
Danny_Larouche earned 250 total points
ID: 16984727
It could be lower but still acceptable for terminal
0
 

Author Comment

by:hrasheed
ID: 16984748
at what point would you begin to notice severely degraded performance?
0
 
LVL 18

Assisted Solution

by:carl_legere
carl_legere earned 250 total points
ID: 16984808
450-500 ms

you should be fine if your ~300ms latency is consistent.
0
 
LVL 3

Expert Comment

by:livedrive777
ID: 16985368
I disagree...  While your users will still function at that level of latency they are going to notice some severe type ahead issues and the like (where they type characters and they don't show up on the screen for several seconds).

Ideally for Terminal Services it is best to keep the latency at or below 100ms, but obviously the lower the better.

Are you using Citrix with Terminal Services?  If so there are a few things you can do to improve the user experience if you cannot improve the latency.

Can you account for what causes the latency?  Is this over the inet, over a WAN connection, VPN, etc???  If the users are inside of the continental US you should be able to achieve much less latency than that if you can pin point where it is being introduced.  Can you post a tracert or something as well?
0
 
LVL 1

Expert Comment

by:rtanner_big_d
ID: 16988001
Dear herasheed,

To assist with livedrive777, are you using Voice-over-IP (VoIP)?  This could account for much of the lattency.  If so, SEPPPPERRRRATTTE thy two systems!!

RT
0
 
LVL 8

Expert Comment

by:Danny_Larouche
ID: 16993692
Livedrive777:  The 300ms is a round trip value,  then the effective latency is less than 150ms.
0
 
LVL 3

Expert Comment

by:livedrive777
ID: 16995553
I could be wrong, and I certainly don't speak on behalf of everyone, but I've never referenced latency in anything other than the round trip value.  That is the value that pings return, that is the value that tracert returns...  In fact you can't track latency from a single point without measuring the entire trip.  What use would it be to know how long it took traffic to get to a destination, but not back??

Regardless of that all of my comments are centered around the idea that the latency is measured based on the RTT.  300ms is too slow, and I would look to get it closer to 100ms or less if you don't want to see type ahead problems and I've been using Citrix and Terminal Services for a LONG LONG time.  This forum recomended a high threshold of 250ms as the HIGH point I'll emphasize before the apps are unusable.  http://support.citrix.com/forums/thread.jspa?messageID=263698&fromSearchPage=true&#263698

I've known of people using Terminal Services over higher latency connections, and everyone's threshhold for pain is different, but I can tell you this...  With the latency you are seeing right now your users will notice and I do not believe they'll be happy about it.

More to the point can the latency you're seeing be resolved?  Over straight INET at least I can get to almost anywhere in under 100ms.  VPNs and such add more overhead, you don't really need to run the Citrix traffic over a VPN.  If I were you I'd be trying to get that latency down to 100ms or less, or at least as close to that as you can.
0
How your wiki can always stay up-to-date

Quip doubles as a “living” wiki and a project management tool that evolves with your organization. As you finish projects in Quip, the work remains, easily accessible to all team members, new and old.
- Increase transparency
- Onboard new hires faster
- Access from mobile/offline

 
LVL 1

Expert Comment

by:rtanner_big_d
ID: 16997389
...the VPN config was only to try and isolate the latency as well as isolating the ports.  How about a possible Loop-Back that needs to be disabled?

RT
0
 
LVL 3

Expert Comment

by:livedrive777
ID: 17345283
Well, I would want to be a pain, and if the poster agrees then no prob, but otherwise I would object to that only b/c Danny_Larouche's comment, "The 300ms is a round trip value,  then the effective latency is less than 150ms" is misleading as I pointed out in the follow up post.  Also, of any of the responses mine were certainly more detailed and I at least cited supporting outside documentation.

Aside from that I still stand by my assertion that the 300ms round trip time is too high for extensive use of Citrix sessions and he needs to get that latency lower.  The only justifiable reason in my opinion for the latency to be that high would be if the users were in India or somewhere overseas.
0
 
LVL 39

Expert Comment

by:redseatechnologies
ID: 17345890
No problems, it is up to the Moderators, not me.  I was expecting this question to draw attention anyway.

I will explain why I made my recommendation.

In my experience with terminal servers, of which, i have many.  450-500ms is workable.  Yes it is slow, but you followed that up by saying "where they type characters and they don't show up on the screen for several seconds"

500ms is not equal to several seconds.

From there, the thread kind of broke down into the definitions of pings and a whole heap of unnecessary stuff.

I awarded Danny_Larouche & carl_legere a split because they answered correctly*, and first (including a follow up from the asker). *correctly, in my experience with TS.

As I do not want to get into an argument (which I am not implying could or would happen) please post if you agree or disagree and your reasons, that way the Mods will know what to do with the question.  If you would like to continue this discussion further, please make a note of that in your post and I will either ask for Mod intervention, and/or create a CS thread for us.  There is enough off-topic stuff in this thread and I would like to keep it to a minimum.

Thanks

-red
EE Cleanup Volunteer
0
 
LVL 3

Expert Comment

by:livedrive777
ID: 17353125
Well, I certainly understand and the definition of acceptable latency can vary widely from person to person, so I can certainly let it rest at that.  I just wanted to make sure those additional factors were weighed in.
0
 

Expert Comment

by:Thirst4Knowledge
ID: 40365989
The interesting thing about this question is that nowhere did the asker state if his users were complaining of poor performance.  The question was only "is this response time slow".  We use citrix yet a branch with twice the latency says the performance is fine than a branch with half the latency
0
 
LVL 25

Expert Comment

by:kieran_b
ID: 40379042
The extraordinarily interesting thing is that you are responding to an 8 year old question :)

It would be reasonable to assume that if users were being asked to check ping times, that someone had complained at some point.  Lots of people post here having already walked down a lot of their own troubleshooting.
0
 

Expert Comment

by:Thirst4Knowledge
ID: 40417257
8 year old question but is still relevant now, hence why as a member I pay to be able to search historical question that may apply to me whether they are 8 minutes old or 8 years old.

 Why have a site to view questions if we are to ignore the answers no matter how old they are. You might as well shut this site down :)
0

Featured Post

What Security Threats Are You Missing?

Enhance your security with threat intelligence from the web. Get trending threat insights on hackers, exploits, and suspicious IP addresses delivered to your inbox with our free Cyber Daily.

Join & Write a Comment

Don’t let your business fall victim to the coming apocalypse – use our Survival Guide for the Fax Apocalypse to identify the risks and signs of zombie fax activities at your business.
Know what services you can and cannot, should and should not combine on your server.
After creating this article (http://www.experts-exchange.com/articles/23699/Setup-Mikrotik-routers-with-OSPF.html), I decided to make a video (no audio) to show you how to configure the routers and run some trace routes and pings between the 7 sites…
In this tutorial you'll learn about bandwidth monitoring with flows and packet sniffing with our network monitoring solution PRTG Network Monitor (https://www.paessler.com/prtg). If you're interested in additional methods for monitoring bandwidt…

746 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question

Need Help in Real-Time?

Connect with top rated Experts

9 Experts available now in Live!

Get 1:1 Help Now