[THIN] Re: Branch Repeater v Riverbeds

  • From: Stuart Thompson <stuthompson@xxxxxxxxxxxxx>
  • To: "thin@xxxxxxxxxxxxx" <thin@xxxxxxxxxxxxx>
  • Date: Tue, 12 Feb 2013 09:09:45 +0000

Thanks for all the input, I really appreciate it. We've certainly been looking 
at the local network also, but so far nothing to suggest concern. I'll be 
revisiting this now. The akamai site looks really interesting, but for 
something else I have in mind! Score - thanks Chris.  Interesting comment from 
Alan too.  We're looking at different configs, looking towards a combination of 
BR using WCCP and only touching Citrix, Riverbed 'doing' everything else.  We 
had limited success on the Riverbeds with optimization policies, so reverted 
those. Ultimately i'm really focused on keyboard and mouse latency as reported 
by the users....   Silly question time again:  Is it somehow possible to 
monitor and report on mouse movement latency?Meaning, other than look and feel, 
how can I say that change X had a negative or positive impact and by what ratio.
Thanks again everyone,StuFrom: andrew.wood@xxxxxxxxxxxxxxxx
Date: Tue, 12 Feb 2013 07:19:13 +0000
Subject: [THIN] Re: Branch Repeater v Riverbeds
To: thin@xxxxxxxxxxxxx

…this is a good call: it’s not just about the pipe between you and them – 
what’s the “local” network like 
 From: thin-bounce@xxxxxxxxxxxxx [mailto:thin-bounce@xxxxxxxxxxxxx] On Behalf 
Of Pat Coughlin

Sent: 11 February 2013 13:39
To: thin@xxxxxxxxxxxxx
Subject: [THIN] Re: Branch Repeater v Riverbeds 
In my experience the "mixed experience with keyboard and mouse" is more related 
to fluctuations in latency.  A rock solid 270ms (India to Ohio) does not create 
this type of issue.  My offshore customers would complain the most when the 
local link began to be hit by other traffic (folks streaming video in the next 
cube).  Despite having an adequate pipe to the US, the link for their office 
would get congested and once Citrix traffic starts getting dumped into the 
buffer at the router performance drops down the toilet.  It is often worse on a 
dedicated link as the VPN hardware will not report the issues on the wire to 
the client and just queue traffic in it's buffer as well.
 Patrick CoughlinCCIA 
On Mon, Feb 11, 2013 at 3:30 AM, Stuart Thompson <stuthompson@xxxxxxxxxxxxx> 
wrote:Hello list!
 
Apologies for not playing an active role in the years i've been following the 
list... however, I turn to you for assistance now and if anyone can offer some 
guidance, i'm sure i'll be able to find it here.

 
ENVIRONMENT:
I'm faced with high latency from Asia region to Central Europe. From 250ms 
upwards. Available bandwidth is currently not an issue.
Users connect to a VDI Windows 7 desktop from Asia to Europe.

 
PROBLEM:
Users have mixed experience with keyboard and mouse latency. (lag, 
slowness...etc etc)
 
BASIC SETUP: (network device)
Currently using Riverbeds v 6.5 and look to v7.0 soon. (v8 out of scope atm)

 
QUESTIONS:
Although i'm not setting out to change physics ;o) i'm really interested to 
know if anyone has used Branch Repeater instead of Riverbed for this kind of 
user experience to any level of success.

 
Also, I am very new to Branch Repeater...  any tips or gotachas for testing 
such a device.  I can replicate the latency with a Network Nightmare device.
 
Any help is greatly appreciated...
Cheers All -  Stu
                                          

Other related posts: