Having attended the previous User Group’s London event in February the Co-operative Group, who are in the process of a 5000 seat Citrix desktop project, kindly opened their swish Manchester boardroom doors to allow the user group to admire great views of a playing-to-the-part rainy Manchester, to share their experiences with a major roll-out, discuss common themes and generally look after everyone marvellously.
Everyone at the Co-op co-operated in making our day great, the receptionist, the caterers and of course the team who made it all happen (thanks Tim Tharratt and David Turton). They weren’t alone – thanks as well to our sponsors AppSense, eG Innovations, RES Software and Whiptail for contributing interesting session and making sure everyone was comfortable and well provisioned at the post-sessions drinks .
Agenda for the day covered
- Co-op: 5000 Seat Citrix Service – Architecture: now for the science part
- Whiptail describing how their silicon storage arrays can make a difference to VDI projects
- 360is walked through a case study on CDT Cambridge and how they went about accommodating difficult workloads in a xenDesktop environment
- AppSense used their sponsor session to talk through: Personalization Best Practices: Tips for Success
- MS-SQL: How important in a Citrix Environment?
- Intro to Netscaler & Netscaler v10: Netscaler, not just for enterprise – it’s a friendly and lovely thing
- XenApp & XenDesktop Best Practices – Citrix’s Andy Baker and Thomas Berger present their Best Practices white paper via Go To Meeting
- eg Innovations described how their monitoring suite can assist a Citrix deployment
- RES Software talked through how to orchestrate and automate Citrix implementations
- Post Synergy Round-up
- [Round Table]: If you had a blank sheet of paper how would you design your Citrix services?
Co-op: 5000 Seat Citrix Service
Tim Tharratt and Neil Burton presented a technical overview of the Co-Op groups branch desktop service roll-out for 5000 seats. The service had to offer secure services supporting isolation of applications, be better able to accommodate change and still improve integration of services.
Using AppSense user virtualisation, Citrix NetScaler SDX, PVS, XenApp, Microsoft App-V and Hyper-V and ThinPrint’s printing solution, the in-house Co-op team were able to virtualise all of their applications for delivery by App-V.
It was an insightful walk-through of considering design challenges and how the different technologies to overcome them. The use of NetScaler SDX for instance to provide a range of NetScaler VPX services to create secure zones for services and lock down application access; the use of App-V to virtualise all the applications led to a change from XenDesktop to XenApp to take advantage of greater user density; XenApp PVS considerations; the need to use thin clients running XPe
An impressive project, well considered with some useful insights. If you get a chance, have a listen to the guys if they present in the future but in the meantime you can download the presentation and have a looksee yourself.
A useful view of the products Whiptail have to offer in relation to improving VDI performance with their optimised storage appliances. Feel free to download the Whiptail slides
Accommodating Difficult Workloads In XenDesktop by 360s
We all know that XenDesktop can be used for knowledge workers – but how about when that Knowledge Worker has a doctorate in Very Complex Stuff, a certificate in Hard Sums and wants to do run data modelling applications that’ll take hours to run? The team from 360is walked through a case study dealing with stressful workloads, options for engaging demanding users and how they helped to update the code to push XenServer further.
AppSense took a different tack with their sponsor slot and handed it over to Matt Murphy to expertly talk about Personalization Best Practices: Tips for Success. Matt talked through the impact that profile size can have on logon times, and considerations for sizing and performance when deploying AppSense Environment manager. Was impressive to hear good critique and recommendations.
Microsoft SQL – How important in a Citrix Environment?
Neil Burton expanded points he’d made in the Co-Op presentation on the importance of the increasing importance of MS-SQL to the smooth running of an enterprise Citrix environment. Sure, if your datastore goes off-line it’s inconvenient, but SQL will often have a far greater role.
It underpins services such as EdgeSight, XenDesktop, PVS – it has configuration information for AppSense and RES environments. The set-up, management and patching of your citrix service databases isn’t always a job that will be undertaken by “the beards” – your DBAs who know SQL in-depth. Neil de-bearded the process, giving useful tips and insights on design, security, backups and maintenance.
You can download the slide deck for further information.
Intro to NetScaler & NetScaler v10 – not just for enterprise – it’s a friendly and lovely thing
I’m going to be honest – I was setting up for the next session and trying to make sure Dave finished on time: I only had half-an-ear Dave Holborn’s presentation. Dave gave what sounded like a very good introduction to the recently released NetScaler v10 and its interaction with XenApp/XenDesktop. He led discussion on the tools that are currently in place, looked at reasons for the continued existence and expectations of what NetScaler could do. He then took a look at the new functions available.
No one threw fruit, or was asleep – and it got great feedback so I’m confident in saying it was very useful.
XenApp & XenDesktop Best Practices
A common theme in feedback forms is “can we have best practices” – Citrix’s Andy Baker and Thomas Berger had recently released/updated Best Practices for XenApp/XenDesktop white-paper. They kindly agreed to present their Top 10 Best Practices via Go To Meeting. Bear in mind Best Practice is right @ 80% of the time – for the other 20% ymmv, but do share your experiences. Their top ten…
#10 Various Artists – use monitoring tools to ensure the service is working as expected; perform capacity planning tests before & during service; implement a multi-stage environment e.g. (dev/UAT/Prod). Develop change management; backup stuff. It seemed as if they’d read 12 Steps to a Better Citrix Farm, but I expect Thomas was just being very clever in his own right.
#9 PVS – be wary of threads and port, because lack of threads being available means a queue, means latency.
#8 Anti-Virus – you’ll have it enabled, manage it – check out CTX127030.
#7 License Server – redundancy and impact in a XenApp PVS environment
#6 Storage – monitor that disk latency
#5 XenServer 6.0.2 – use processors that support SLAT, disable Green IT Bios functions; consider your DOM0 RAM size
#4 XervServer – think about pool design with resource pools for Control Infrastructure, XenDesktop, XenApp
#3 Processor & Memory Overcommit: When its sensible, and when it isn’t
#2 Hyperthreading: Not always worth the hype
#1 Scale Up or Scale Out? There’s a lot to consider isn’t there?
Given the impressive AV facilities in the room it was slightly disheartening to be let down for a minute or two by internet audio deciding it fancied a wee, but still a useful way of sharing information and direct access to a wider Citrix resource. Perhaps next time we’ll look to widen the audience by opening up GtM sessions to a wider audience and not just those who sent texts/tweets asking if they could have the link.
the egInnovations team highlighted Thomas’ point of having good service monitoring with the consideration that Edgesight is currently being shown as being EoL in 2013 (this has since been confirmed as being incorrect – EoL will be in 2018) to highlight not only the Citrix monitoring capabilities that egInnovations can offer but the monitoring of other services too giving a wider view and understanding of performance issues and trends. Download the Best Practices For Citrix Performance Management slidedeck
David Hitchen took a refreshing tack on the RES message and queried why anyone would want to install XenApp and XenDesktop… manually. An adriot introduction to two of RES’s six (yes, six) products. David highlighted how Automation Manager and Service Orchestration, again with reference to the best practices discussed by Andy and Thomas, can be used to help deliver a reliable, stable environment – and showing that it’s not always a case of either RES or AppSense.
Post Synergy Round-Up
We had a chat around Podio – which was generally seen by those who had had a look at it as a useful, easy to understand web based tool for collaboration. Your first thought may be “it’s a Sharepoint Lite”: the discussion was around how the framwork was far easier for users to get to grips with. It may be useful to note the the user group is currently understanding what is involved creating a shared resource environment to manage the user group. We’ll update you on that as we go.
XenClient Enterprise – for those who had looked at XenClient prior to Citrix’s acquisition of NxTop, perhaps time to look again. Nxtop is a very different product, with a far wider HCL. There was discussion around its place within a XenDesktop environment, whether it should be offered standalone. If it was still relevant to have a client hypervisor solution given Windows 8 has one built-in and what the impact of Citrix Remote PC would and could be.
Other topics included Project Avalon (very exciting, the convergence of products coming together); updates to CloudGateway, and new features in ShareFile (e.g. on premise/off premise, integration with other services).
There was likely more detail, but at this point I was sat here X …………………………………………………and my pen and paper were over here Y. It was a very interesting discussion tho’, likely to be followed up in discussions and wound it’s way neatly to..
Roundtable: If you had a blank sheet of paper how would you design your Citrix services?
The roundtable is always a popular session – key discussion points included.
- How lucky were the co-op? I wish I could virtualise all my Apps! Application rationalisation was seen as key to a successful project. A key component of the Co-Ops decision to use XenApp was the fact that all apps could be virtualised – so each build OS was extremely clean. That said, it was agreed that didn’t detract from the fact that the Co-Op architecture presented was well designed, well considered and an exemplar for how services should be put together.
- ThinApp, App-V, Citrix Streaming – which one? (again) A similar question to that in Feb. Again, experience suggested that Thinapp is the easier to work with, and has a range of deployment options, but has little in the way of deployment management options unless you choose .msi files and integrate with AD. App-V was suggested as the most comprehensive solution available: but can be more taxing in terms of initial effort and resources required. Likewise, there was a thought that, if you’ve selected Citrix Streaming, go you, everyone is entitled to their own opinion. Still, App-V often wins out because of its breadth of application support, range of management options and the fact that for many customer you may well already be licensed to use it across a range of environments – with the latest version giving even greater benefits.
- Include Monitoring and Orchestration: quite often the ability to orchestrate and automate provides great benefits for stability and the ability to adapt to change. The enhancements in newer versions of XenApp are very welcome. It was interesting to see the RES and egInnovations offerings and be able to relate that to the best practices that Andy and Thomas provided.
- To never have to print, ever. The co-op had chosen ThinPrint: branches still need to print. There was general discussion on the benefits and trials of such solutions – bandwidth, driver-less printing, accommodating legacy devices: we’ll likely make a number of sessions on these in the future. The key, don’t underestimate an organisations need to print, or the user experience when printing – design early to accommodate, don’t tack it on at the end.
- More NetScaler To be used more, and not just as an edge device, but as a method of providing separation between services, load balancing and higher availability.
Again, thanks to AppSense, eG Innovations, RES Software and Whiptail for sponsoring the event, providing drinks and platters, and a very honourable mention to the Co-Op group for ably demonstrating northern warm and charm.
Next user group meeting will be likely around Q3ish, venue(s) will definitely be in the UK – so Monaco very likely to out I’m afraid David. It’s a goal of the user group to share ideas and information. If you see a pertinent presentation, if you plug into wonderful webinar if you’ve done something you simply want to share then the UK Citrix User Group exists to help you distribute that resource, indeed enhance it by having a place to discuss it with your peers.
Join us on Twitter where our lists now includes links to all those who have presented.
Share with us at up and coming events, or on our LinkedIn group