citrix related material
So I’ve been with my new company Liquidware Labs a few days now and I think it’s about time I explain what we do if you don’t know us. Plain and Simple we enable VDI. Well what does that mean? I think until recent time it’s been quite difficult to truly assess and design a VDI deployment and then if you wanted to migrate and manage VDI you end up with a lot of moving parts. Too many variables have to be accounted for and trying to manually assess and design is a nightmare. There are some great references on the blogosphere like Andre Leibovici’s VDI calculator found >here< but even this assumes you know things like how many users you have and how each one is using a static amount of resources. So how would you get that information in the first place? Don’t get me wrong if you have nothing else to go on Andre Leibovici’s VDI calculator is a work of art. Enter Liquidware Labs. We provide full visibility of what’s going on with your physical world and help you understand how to design a VDI deployment. With StratusphereFIT we analyse key metrics and not just CPU, Memory and Disk.. We look at Logon times, application response times, whether applications are good candidates for VDI and a plethora of other stuff that is truly required to understand a deployment going forward. Sorry Andre but whilst I love your VDI calculator and again stress it is worth creating these free community tools, there is no need for it with StratusphereFIT. FIT kicks out several detailed reports and highlights which Users, Apps and desktops are good/fair/poor candidates for VDI. You can then manipulate this information and design a VDI deployment from it.
If you know me you would know I hate to sell the future. I only ever do this if I know a product/feature is literary on our doorsteps. Coming very soon Liquidware Labs will release a new component known as StratusphereDISGNER. Before with FIT we provided full visibility and with some guidance expected you to go about designing your deployment yourself. Now with Designer we are even going to help you to craft the design as well.
So what happens next? Well if you are a consultant what would normally happen with this kind of information to hand you perform a POC and if it works job done. Or is it? Well the feedback we have is that a lot of VDI projects stall because incorrect methodology has been used for the lifecycle of the VDI deployment. This is a real-world scenario I keep hearing similar stories to; consultant deploys VDI, initial POC proves good feedback, POC goes Production, feedback is good. The end-user then adds a few more users to the system and performs drops. End-user is not happy because requirement was supposed to be future proofed. Typically what happens is the storage platform gets the blame. And sometimes this is the case. But how do you know and should you trust your consultant when he wants you to throw even more money at fixing the problem? Well enter StratusphereUX. UX uses the same metrics as before used in FIT but now instead of reporting capacity UX will help you pin point bottlenecks. If it is your storage platform, fine…but what if it’s a rogue application that is hammering the storage that doesn’t need to be there? Then you may have thrown money at the problem only to find out it’s not fixed the problem. To protect yourself you need visibility of all key metrics down to user and application.
Question: Can I use a standard infrastructure monitoring tool or virtual infrastructure monitoring tool to give me the same information.
Answer: Probably not. This kind of tool most probably doesn’t have visibility of key VDI metrics like Logon times, application response times and even how graphically intense an application is.
Question: What using a VDI specific tool like Xangati?
Answer: While Xangati has a very nice looking GUI and has visibility of VDI protocols like ICA or PCoIP it is still missing those essential metrics: Logon times, application response times and even how graphically intense an application is. These are vital for understanding user experience.
We have clients that have told us the ROI is high when using UX. The answer to this is simple: do you want to spend £60k on a new storage unit or find out actually it’s just a buggy application 😉
What if a user complains his/hers virtual desktop is slow compared to when he/she had a physical desktop? Wouldn’t it be great if you could demonstrate there was no difference other than perception? Well you can with StratusphereUX
What I have described up to now is stages 1, 2 and 4 stage of a VDI deployment lifecycle. Assessment, Design and Validation. But what about getting the users away from physical desktops on to VDI? Enter Liquidware Labs Profile Unity. Profile Unity is a top class (cheap) Persona Management solution that is key to the migration of users to and from VDI. As well as the usual profile management stuff you’d expect to manage the user environment it can also migrate the existing settings and data from the original desktop into the VDI world. Why is this important well to me I remember back in the day when I was a Citrix consultant we would have to perform migration manually. So this feature of Profile Unity is a major time saving feature. You noticed I used the word “cheap” before which is unusual in this industry. No one likes to buy cheap software for production, right? Well actually this is a very important statement because the problem with VDI is it’s not cheap. If you are thinking you are going to save oodles of money on a VDI CAPEX compared to using physical desktops think again. RIO comes down the line when you appreciate cost saved in time of management but initial outlay is high. So my point about cheap! Anything to drive down the CAPEX has to be a bonus in this day and age. Migration is the third stage of a successful VDI deployment.
Now for some more good news. These 4 solutions can be deployed as a single virtual appliance known as the Stratusphere HUB reducing the need for tons of backend infrastructure. The picture below demonstrates the architecture.
Notice we deploy an agent? Well the only way to pull the essential metrics we need is by using an agent at the endpoint. That is probably why products like Xangati don’t have all the vital metrics.
One last word from me: Next year is the “year of the dragon” but it’s also the “year of virtual desktop”
So I have some big news if you’ve not already heard. I’m moving on from Veeam Software where I was the Director of Global Education Services to join Liquidware Labs as Northern EMEA Technical Director. I left Veeam with my head held high whilst staying friends with the Veeam guys. Veeam provided me with an opportunity to show off my talent in vendor land after spending many previous years in the channel. My contributions to Veeam in the 3+ years I was there included:
- Helped onboard a ton of Partners in Northern EMEA
- Authored several whitepapers/techpapers
- The formation of the Education Services team
- The ProPartner Accreditation program to train partners on sales and technical education of which Veeam nearly have 12,000 people enrolled. This I was my proudest moment as my team and I received tons of great feedback about this project. Some comments included that it was remarkable that a vendor the size of Veeam could pull off such a program. Even one of our competitor admitted they’d failed at trying to get the same thing off the ground.
- The new End-user training program which will go live after I left
As you can see moving wasn’t a decision made easily but my reason is simple. I enjoyed my time at the beginning when Veeam was in start-up phase and I was creating a buzz by getting partners hyped up about Veeam. An opportunity came my way to do that all over again with Liquidware Labs and to top of that Liquidware Labs are VDI focussed. You may not know this about me but before joining Veeam I saw myself as a bit of an expert in VDI. For me it was a logical progression. I’d worked with Citrix/Terminal Server/Server based computing for many years in a previous life. So when the concept of virtual desktops started to rear its head in 2006 it clicked with me straight away. I even have notches on my belt like I was the first VCI outside of VMware to run a VMware View course (when it was still called VDM) in EMEA. So my heart was always with VDI and I lost my way for a few years whilst I immersed myself with Veeam products but now I’m back. Back to Liquidware Labs; there are tons of start-up vendors I could have approached but I only paid attention to those that the community is hyped about. Liquidware Labs has a good vibe and great reputation. I feel the same buzz about them as I did with Veeam at the beginning.
As you can imagine my blog will change direction to showcase the value I can bring to VDI and Liquidware Labs solutions. If you are not familiar with Liquidware in brief terms we bring solutions that address the following key areas in VDI deployments:
- Assess – finding out how your virtual desktop environments look like now and understand what can be virtualised and what cannot be virtualised
- Design – recommend a design (Hardware and Software) from key metrics that are most important when designing a VDI solutions
- Migrate – transparently migrate users from their physical desktops whilst ensuring the user experience isn’t degraded
- Validate – enhance the user experience after migration by monitoring key metrics and understanding VDI bottlenecks and issues
As for me I believe that is the right time for me to jump back into VDI. 2007 I presented on tour on a VDI road show in the UK and back then people were just not sold on the idea of VDI. Technology wasn’t there yet, cost of infrastructure or lack of education whatever the excuse was, adoption was low. Since then every year I hear VMware say this is the year for VDI but actually I really believe this is the year for VDI. I based that on activity with partners. Plenty of friends and acquaintances from the partner community are telling they cannot get enough VDI nowadays…
That’s enough for now but watch this space.
If you have any Liquidware Labs questions that you want my help on my new email address is email@example.com
Back in 2006 I wrote whitepaper outlining my thoughts about running Citrix XenApp virtualised in a ESX VM. The information is probably still relevant so if you want to read it Click Here > Running XenAPP in vSphere (5706)
Recently I had a conversation with someone from VMware and we agreed that: there’s no need to remove vmmemctl, as installing it will not have any negative impact on the VMs by itself. Even if you do not plan to over-provision your ESX hosts in terms of memory, it is in your best interest to have the Balloon Driver as an insurance should memory pressure increase (i.e. because of a host failure and HA restarting VMs on the remaining hosts). If vmmemctl is not installed, ESX will have no other way than to reclaim memory by swapping out Virtual Machine memory, which will be a more severe performance impact. Here a somewhat recent post by Scott Drummonds on that topic: http://vpivot.com/2009/09/25/esx-memory-management-ballooning-rules/”