Subscribe RSS Feed | Follow on Twitter

You want to tell me PCoIP performance is bad?

with 7 comments

Update: I want to clarify that I of course know that network connections all over the world are different in speed, latency and packet loss. My video only shows that my ad-hoc connection from Europe to the West Coast just works well and the video/website mentioned in this article could give the readers/watchers/listeners are wrong impression of PCoIP.

This week I was really surprised about an blog article I saw linked on Twitter. In the article/videos the author shows a virtual desktop connected over PCoIP from San Francisco to Ottawa. For more details on the scenario please check the article and also listen to the speaker in video. He said: “Definitely much much worse than ICA was…”. Well the “real world” scenario as he describes it looks really strange to me. Ok, he’s connecting from a hotel internet connection (which is mostly crap, but in SF?) via VPN to his data centre but only on one continent with a distance of approximately 5100km (Maybe the cable is going a longer way…). I know PCoIP and is has definitely a much better performance on the WAN and I checked it. This is my real world example: My connection is from my home office’s DSL line through the VPN to the data centre in California.

image

A ping from my local workstation to the View Manager Server looks like that: (Doh! 382ms)

64 bytes from xx.xxx.xx.xx: icmp_seq=2 ttl=242 time=382.313 ms
64 bytes from xx.xxx.xx.xx: icmp_seq=3 ttl=242 time=382.777 ms
64 bytes from xx.xxx.xx.xx: icmp_seq=4 ttl=242 time=382.545 ms
64 bytes from xx.xxx.xx.xx: icmp_seq=5 ttl=242 time=382.215 ms

And the result looks like that:

I think this is great for an overseas connection and watching a video, hah? Sometimes it hangs for a short time but 382ms!

Similar Posts:

Written by Christoph Harding

May 21st, 2010 at 5:14 pm

Posted in PCoIP,VMware View

Tagged with , , ,

«
»
  • Anon

    Awesome! What would happen to your connection with 2, 3, 4% loss? Not everyone in the world has access to the exact same conditions, scenarios and environments. With no loss any protocol is indeed awesome. Add loss to the picture and all suffer, at different degrees.

  • http://www.cd-k.de/ Christoph Harding

    I love the anonymus comments most. How many loss does my connection have?

    Von meinem iPad gesendet

    Am 22.05.2010 um 04:09 schrieb “Disqus” <>:

  • Shanetech

    Thank you for this post. Just for some grounding, I support a 2000 User VMware VDI (pre-View) 2.1 / View 4.0 environment and a 3500 CCU Citrix XenApp / XenDesktop environment. These environments support 240 sites on 4 Continents. Our WAN infrastructure runs the gamut from a 2Gbit Internet pipe in the US HQ with a 45Mbit redundant MPLS pipe to EU HQ all the way to 56Kbps line of site connections in hostile areas like Afghanistan. My point is and what the comment before me was alluding to, is that many WAN Infrastructures are not predictable and this connection, albeit bordering on high latency, is consistently latent in a range within 0.562ms (according to your ping test which is not the complete picture of this actual connection). This level of latency is the norm and the actual SLA from many providers for connectivity to countries like India or China from the US and EU. I concur with the user experience posted and have also experienced similar results given similar conditions. However, I feel your post lacks the specificity required to truly debunk the link posted in your article. I feel your demo, instead of trying to negate another post, should simply show your experience on a connection with an average latency of 382.4625ms.
    In the spirit of professionalism and technical accuracy, would it be possible for you to update your post with granularity and depth in the following key areas:
    Detailed definition and explanation of the actual WAN connection
    Leverage other tools besides ping to gather connection metrics (many tools are available, these tools are especially common in the VoIP arena)
    Possible introduction of Packet Loss and Jitter via a simulator
    Show RDP, ICA, & PCoIP all under the same conditions

    In summary, I feel the community should work together on providing the laymen and the technical savvy with quantifiable and detailed output. I, like many, love and post YouTube demo videos of technology in this space. However, once we begin to compare / debunk scenarios and technology, especially remote display protocols and networks, we owe it to the community to be as thorough and granular as possible (Just my opinion). Lastly, I stayed at the same Hotel at the same time as the poster of the first video. I connected to my VMware View / Citrix XenApp infrastructures from the Hotel also. I guess due to Citrix Synergy and the demographic present (everyone has a laptop and was connected) the experience wasn’t the best. I experienced variable latency from 102ms to 1200ms and packet loss to a data center in New Jersey. I dropped my sessions via ICA, RDP, VPN (and PCoIP). At one point, I switched to 3G and stuck to Web Apps / ICA apps only.

    I look forward to your next post and thanks for taking the time to create a View centric site. The community certainly needs more of them to balance the already established and many Citrix blogs that currently exist.

    Take care,

    Shanetech

  • http://ag4it.myopenid.com/ AG4IT

    PCoIP is certainly a great display protocol. However in some scenarios of slow remote connections (like over certain WANs) there may be issues where PCoIP doesn't function quite as well. In those cases, you can complement the VMware View deployment with Ericom Blaze, a software-based RDP acceleration and compression product that provides improved performance over WANs. Besides delivering higher frame rates and reducing screen freezes and choppiness, Blaze accelerates RDP performance by up to 10-25 times, while significantly reducing network bandwidth consumption over low-bandwidth/high latency connections.

    You can use VMware View with PCoIP for your LAN and fast WAN users, and at the same time use VMware View with Blaze over RDP for your slow WAN users. This combined solution can provide enhanced performance in both types of environments, letting you get the best out of VMware View for your users.

    Read more about Blaze and download a free evaluation at:
    http://www.ericom.com/ericom_blaze.asp?URL_ID=708

    Adam

  • Tim

    I am wondering if anyone is having the same issue as I am with audio quality while using PCoIP.
    I am on View 4.0.1 running on ESX 4. I have a great deal of abnormal clicking, popping and echoing while playing a simple wav file. The simplest test can cause this. For example, If I play the default beep from the control panel more then once the sound is very bad. The first beep may be fine, but playing it repeatedly causes terrible distortion.
    It would be great if someone could try that test on their setup and let me know if they get the same distortion.

  • Mark D.

    Interesting read and thanks for posting.  I bumped into this article while doing some
    research on this exact same subject. Not sure what tools everybody is using,
    but we have used the free software here http://www.cibengineering.com/blog/pcoip-packet-loss-latency-detection-software/
    to help determine where packet loss is and further diagnose latency.

  • bw

    Why does everybody use video in all their testing? Are business users all watching people water ski and playing video games? Or are they looking at spreadsheets and working in Outlook?