Wilco van Bragt - LinkeIn Wilco van Bragt - Twitter rssa 

E2E VC London 2011 Day 2

Day two of E2E VC started for me with a presentation of Aidan Finn about Hyper-V in Windows 8. He started with the increased scalability (cluster/host/VM maximum possibilities), followed by the features (hyper-v replica [disaster discovery], bitlocker and HA, Live Migration separated from Failover Clustering, RemoteFX via WAN, Virtual fibre channel adapter, guest NUMA). Aidan continued in more detail about this feature, starting with Hyper-V replica.

Check the slide for all information, too much to write down. Next topic was Bitlocker for Hyper-V (everything can be encrypted now, including the shared storage volume). Aidan continued with Virtual Networking, now NIC teaming is supported, extensible switch (partner expandable), network virtualization, Private VLAN,  PortACLs and QoS are keywords about this topic. Network virtualization: run multiple virtual network on a physical network, each virtual network has illusion it is running physical. Lots of improvements made on the network level. NIC teaming is not especially for Hyper-V, it is also available for physical Windows 8 machines. SR-IOV is explained in detail, the network is directly connected to the VM, wither the hypervisor in between. Features for clusters are: guest application health monitoring, VM failover prioritization, drain all VMs from a clustered host, cluster-aware updating, Next topic was storage: Remote Direct Memory Access, storing running VMs on SMB 2.2, storage spaces/storage pools, VSS provider for VMs on file servers, Active/Active File Server Clusters, Live Storage Migration, Dedup/Thin Provisioning and online disk repair. On the management suite there are 150 built-in PowerShell cmdlets, charge back resource usage metrics and snapshots (live AVHD merge).




Next session was Performance and availability networks for Hyper-V Clusters by Didier van Hoye. Didier was explaining that Jumbo Frames that improves live migration speeds by 20% and less CPU usage (working best for large memory VMs). Power Configuration  should be set to Maximum performance on the hardware level. Next topic was Network Teaming Benefits (throughput/fault tolerance) and Challenges (no official MS support, network issues). Didier discusses several high speed network features, but that was a bit too network technical for me. Next topic was high availability and performance with considerations like number of NICs, rackserver versus bladeserver, converged networks / specialized networks, physical versus logical separation, ROI/TCO and complexity.

Next session I attended Chris Marks session about application delivery strategy. He started with stepping the customers mind and what they are thinking and how a reseller reacts on that. Chris point out his vision with a sheet that are the assumptions the use case will be, but that the approach transaction should be with open eyes (and not about IT resources, but about the processes, users and application owners and how to can access the data they need). Also Chris mentioning the opportunity to use local fat client as a solution. Chris continued with Application assessment and compatibility and the challenges in that area. Chris hits an important point: user segmentation (define core apps, segmentation options: user location, application performance, endpoint required, risk profile).

Personalization - Speed Dating Style by Kristian Vidal was my next session. Via the Speed Dating mechanism he explains how User Personalization of AppSense works. A nice starting point to explain the technique. Recommend checking the slide/videos for the details if you were not at this session. After the technique is explained the session continued with new features in version 8.2 (released this week). Folder Management (change copy, mirror copy, synchronise copy), powershell support, API support policy (commands without needing the GUI), Import/Exports EMPS, SQL mirror support, If Else condition and a Browser Interface (for 1st line support) with good features.

After lunch I attended Andrew Wood's session Citrix XenClient plus comparison with MokaFive and NXTop.  Andrew start with the statement that this session is not another smackdown. He started of how Citrix are positioning XenClient and his vision what the purpose is of a hypervisor (more images and/or an common image). He continued with the type 2 Hypervisors - require a base OS and type 1 hypervisors - installed directly. Andrews continued with client hypervisor type works for your organization and how to compare those hypervisors.  Create a list of requirements, with topics like hardware support, User Experience, how to manage and create VMs and manage them. Andrew compares them using a British card game called Top Trumps (cool idea). Special note to Mokafive not to share a free license to test the product. NxTop scores high on management and control (but should be better positioned by the vendor), according to Andrew. The session ended with Andrew's thoughts of the future of hypervisors (part of the processor) but  the question stays to manage that part.

Just before my session Nando Lutgerink also had a 900 second session about Knowing what to migrate. The session was about the service RES is offering to find out what's being used on the current infrastructure. RES solutions runs as free online service in the cloud (RES Baseline Analyzer). Locally logfiles will be created , which needs to be uploaded to the service again for creating the reports. Nando explains what's needed to get it up and running (for example using logon scripts or via an MSI). Six reports are created (hardware inventory, aplicaiton landscape, user population, location layout, printing topology, analysis summary).  The output files can be used by the Workspace Designer.

Then my first session was on dealing with restarting XenApp 6.5 with using Load Evaluators. I hoped the people who attended enjoyed it.

Roy Textor was next after my session with the presentation titles XenServer & PVS - Case Study: Implementing XenApp in months or minutes. The presentation was a real case scenario for six hospitals in Germany. He showed the current situation, followed by the requirement and the steps within the project (sizing, getting resources and the schedule used). Roy continued with showing the current infrastructure. Actually I expected another content for this session more about XenServer and PVS challenges and experiences.

Before I was the last one for day Mark Knousse presented XenDesktop 5.5 versus VMware View. He created some cool t-shirts, which he gave away. He started with choosing the hypervisor (when using XenDesktop) and really consider the back-end resources needed (like SQL, Licence Servers and so on).  Followed by the infrastructure for View. Next Mark showed the features available in the several available XenDesktop editions and his opinion about this options.  Again followed by the features of VMware View.  Next topic was comparing the installation components, database requirements, support virtual infrastructures, management, logging, client connection software of both products. Mark continued with Aero Glass support and what's required to have to enroll it for both products again. Also the available User Profile solutions were discussed. Mark got into detail about the VMware Linked Clone. Last topics were printing and application delivery compared. The session ended with Mark his experiences.  

I was last this day with my presentation about PDI: Physical Desktop Infrastructure. For those attended I hope that you enjoyed it.