Since the agent offers a silent installation, I thought it would be best to let SCCM do what it does best. I created a package of the agent with a program “DPMAgentInstaller_x64.exe /q dpmservername“. Unfortunately the advertisement fails with a failure exit code of -2147023288.

A quick search didn’t give me any useful information, but since the command itself worked when run from a command line, I decided to try a task sequence and first create the appropriate firewall rules. This should be done by the installer when using the /q dpmservername but I thought it wouldn’t hurt.

The task sequence has 5 steps, the first 4 steps are “run command line” steps, the last one is “Install software” and is the same program that keeps failing when run without a task sequence.

netsh advfirewall firewall set rule group=\”@FirewallAPI.dll,-29502\” new enable=yes

netsh advfirewall firewall set rule group=\”@FirewallAPI.dll,-34251\” new enable=yes

netsh advfirewall firewall add rule name=dpmra dir=in program=\”%PROGRAMFILES%\\Microsoft Data Protection Manager\\DPM\\bin\\DPMRA.exe\” profile=Any action=allow

netsh advfirewall firewall add rule name=DPMRA_DCOM_135 dir=in action=allow protocol=TCP localport=135 profile=Any

After creating the task sequence I assigned it to some test machines and they all completed the task sequence successfully.

Now it’s just a matter of assigning them within DPM to a protection group.

Congratulations! After connecting your mobile device to Exchange Server 2010, you can now send text messages from Outlook 2010 and Outlook Web App on your computer. You can also receive and reply to text messages.

But he didn’t receive any mails or calendar items.

During my investigation I logged on to the Exchange 2010 server and run the Powershell command Get-ActiveSyncDeviceStatistics –mailbox [user]. In the list of configured devices that are able to synchronize with the Exchange server I saw his old Windows Mobile phone and his Android phone. For the Android phone the LastPolicyUpdateTime and LastSuccssSync fields where empty.

I removed his mobile devices from the ActiveSync by using the command Remove-ActiveSyncDevice –Identity [id], (probably the following will also work: Get-ActiveSyncDeviceStatistics –mailbox [user] | Remove-ActiveSyncDevice, or definitely Get-ActiveSyncDeviceStatistics -mailbox [user] | %{ Remove-ActiveSyncDevice –Identity $_.Identity}). After I removed the devices he was able to successfully connect his phone to the Exchange server.

Users can also remove unused Mobile Devices themselves by using OWA. Going to Options -> All Options -> Phone you can see all the registered mobile devices connected to the mailbox.
What I noticed is that since Exchange 2010 SP1 the amount of mobile devices that can be synchronized with the mailbox is limited to 10. When this amount is reached no new devices can be connected to the mailbox and old devices need to be removed.

I think this is something to keep in mind when users complain they can’t connect their new phone to the Exchange server. The amount of devices can be changed using the Set-ThrottlingPolicy command.

A quick look on our UAG server learned that the DirectAccess Monitor reported the DNS64 as Not Healthy.

Looking through the UAG services I discovered that the Microsoft Forefront UAG DNS64 Service was not running and had the start type Manual. Starting the service restored the corporate connectivity for the clients. I’ve changed the startup type to automatic to make sure DirectAccess continues to work after a reboot of the server.

I tried to import the drivers manually instead of using the Dell tools and got errors for a lot of drivers: ‘the selected driver is not applicable to any platforms’

There is a hotfix available that allows driver importing for WinPE that are only signed for Windows 7 at http://support.microsoft.com/kb/978754 (KB978754).

It seems you have to request and install the following hotfix: KB2412273

The added support for the Exchange 2010 online archive isn’t mentioned in the release notes, but trust me… it’s there!

The past four days at TechEd I’ve brought my laptop with me. The first two days I’ve tried my Dell Precision M6500, which is way too heavy and lasts on the battery for about 5 seconds… no, that’s not true, but I struggle to get 2 hours out of it. On top of that it’s too heavy and big to have on your lap to type while you listen.

The second two days I’ve brought my Dell Latitude XT2. Which is way better in terms of battery life (close to 5 hours on the standard battery!) and size and weight. But… of course there is a but… note taking is excellent, the convertible screen is really brilliant, but I don’t like that websites don’t fit on the screen because of the crappy resolution. When in portrait mode you have to set the zoom level in IE at 75% otherwise even the most basic website simply doesn’t fit.

Apart from that… being used to a laptop with a resolution of 1920×1200 and a Core I7… anything that’s substantially slower tends to annoy you a bit. And a 1,4Ghz Core 2 duo is not to slow for Windows 7 and Office, but in comparison… the difference in performance is just too big to be comfortable.

I can hear you thinking… “What about an Ipad?” Well, for starters, it lacks both input methods that I personally really like. A pen is ideal, if the Ipad would support that I would defiantly try it. But for now I think I’ll just dream about an 8″ device, running Windows 7, comes with both capacitive and digitizer support… and has a battery life of 8 hours. All of this of course while delivering the same experience as on my every day laptop… or at least close I know that it something that doesn’t exist… but hey… I’m allowed to dream right? J

Tomorrow the last day of TechEd Europe… looking forward to a session about Direct Access!

Today was the first time that the name System Center Configuration Manager vNext wasn’t used anymore. It’s announced that the next installment of SCCM will be called SCCM 2012. So yes, unfortunately we have to wait a bit longer before we can see it in a production environment.

New features;

SCCM 2012 will have a new architecture, more on this later. It also will have some new features;

Remote Control

Remote Control is back! SMS 2003 was the last version that had this capability, it was dropped in SCCM 2007 because in certain situations it would leave a system unlocked with an administrative account logged on. In SCCM 2012 it will return with some enhancements to make it more secure.

Roll Based Access Control

Part of the new architecture is RBAC. It’s now possible to give access to a limited part of the interface (and functionality). For instance, just the patch management. Or just the reporting part.

Mobile Device Management

System Center Mobile Device Manager will be integrated with SCCM 2012. It will integrate with Exchange to make it possible to manage the Exchange features from SCCM. And it won’t just support Windows Mobile. Nokia is fully supported and there is also limited support for IOS4 and Android!

Patch management

Offcourse, patchmanagement in SCCM isn’t new. What is new is that you won’t get RSI anymore from approving updates! There are now automatic approving policies. You can define policies in about the same way that you make rules in outlook. You can choose to apply them on workstations or servers or both and you can schedule how often it should check compliance. That way you can make sure Forefront gets updated every day, but the rest of the updates only come once a month.

OSD

One of the greatest features of SCCM has been the deployment of systems. In SCCM 2012 there will be support for offline maintenance to your image. You can let SCCM automatically mount your image in lets say the weekend, apply all approved updates and unmount the image. This way your image will always stay compliant and it will save time in the deployment process itself.

New Architecture some changes

SMS has always been about the systems. Not that surprising since it’s called System Management Server… but the world has moved on and now we’re challenged to manage users that have multiple devices to perform their daily work on. That’s why SCCM 2012 has been made User Centric. It’s now possible to assign software to users and define policies so SCCM will act different if you log on to your corporate desktop compared to login on from your personal laptop.

No, your collections and advertisements won’t be a waiste of all your time, SCCM 2012 will still support those, but it’s much easier now to deploy a new version of a software package. If for instance Adobe Reader gets upgraded, you can now say it should supersede older versions. If a machine receives an advertisement for an older version, SCCM will automatically decide to deploy the latest version instead.

Another nice new feature is that you now have option to enforce the desired state of a machine. You can define a policy so that once a machine drifts from the desired state, it will automatically gets remediated. Great if you users have admin rights and tent to destroy office for instance… Office will get automatically repaired.

I can’t wait to get this going at Andarr!

What I would like to share from today is mostly about Lync Server 2010 and a bit about SCVMM.

SCVMM

An upcoming change for SCVMM is the addition of a new feature called Rapid Provisioning. Storage vendors like EMC have been working on this together with Microsoft to help enhance the performance of live migrations on Hyper-V. Especially when using Cluster Shared Storage.

Until now a migration, or deploying a VM from a template would be done over the network, basically making your network infrastructure the limiting factor. This is also one of those “features” (or lack of them) that still make VMware ESX the more obvious choice for virtualizing your highly available servers.

As an example the following case has been presented by EMC:
Deploying 1024 virtual machines, with each a base disk of 50GB, would take approximately 52 hours when using a normal setup with SCVMM.
Deploying the same 1024 virtual machines, but using SCVMM with Rapid Provisioning it would take 1,5 hours! Mind you, I can’t think of a situation where I would feel the need to create 1024 virtual machines before lunchtime… but it does effectively show how much time can be saved when you let your expensive storage solution do the hard work for you.

Together with features like Dynamic Memory Hyper-V is getting more and more a more serious alternative for VMware vSphere.

Lync Server 2010

Lync has some serious changes when comparing it to OCS 2007 R2. Whereas OCS 2007 used to store settings in Active Directory, its own SQL database and WMI, Lync uses a SQL express database on every Lync Server. It still uses AD, but only to query and all settings are stored in its own database.

Another change is the way you upgrade from OCS 2007 to Lync. In the past the client side got updated and you worked your way from there. Now it’s the other way round. When introducing a Lync 2010 front-end server in your infrastructure, your users can connect with the ‘old’ clients and still use the edge server from the OCS 2007 deployment. When successfully tested, you can then replace all other OCS roles.

The last nice thing; all Lync Server 2010 roles are supported when virtualized. Not only is it no longer required to use a physical machine for your mediation server. But it’s also supported to co-locate the mediation server role with the front end server. Making it a more practical proposition for smaller companies that don’t make lots of phone calls.

During this session Brad Anderson welcomed everybody and explained Microsoft’s vision. A lot of that vision is about the cloud and how the cloud can enhance the user experience.

SP1 for Windows 7 and 2008 R2

One of the new features in SP1 for Windows 7 and 2008 R2 is Remote FX. This brings the rich user interface that Windows 7 has to offer to the remote desktop. The workload for this is done by the GPU in the server, so no extra workload is added to the CPU.

Another new feature in SP1 is Dynamic Memory. VMware is already providing this in vSphere, but now Hyper-V will also benefit from a better use of memory. Dynamic memory will help Hyper-V to host 40% more load on the same hardware you are using today.

The Cloud

Microsoft’s vision on cloud computing is a user centric one. In an ideal world it’s not important how, or with what device, a user connects to the information in the cloud. Even the choice of the type of cloud (private, public, hybrid) should not make any difference in the end users experience.

To explain this a brief demonstration has been given with Windows Phone 7. Several applications were shown and explained how the information in the cloud is transformed into a uniform interface experience. Whether you are browsing eBay, or looking for a movie, the interface and the way you interact with the information stays the same.

Microsoft believes that we’re only at the start of the ‘cloud-age’ and that the endpoint should be Platform as a service. Software as a service and infrastructure as a service is only the beginning.

Microsoft’s offerings in cloud services have all been shortly demoed.

Microsoft Azure
Office 365 (formerly known as BPOS)
Windows Intune
Hyper-V
System Center

With System Center Microsoft is working hard on a single management console for all of your services. Whether it’s running locally in your own datacenter, or on Microsoft Azure. You manage it from the same console. Personaly, I like this much better than a web interface that ‘almost’ gives you what you want. A good example of this is the web client from vSphere. It works… but the vSphere client is so much nicer to work with.

See you tomorrow for information from day 2!

After correcting the settings on the cluster it showed on the summary page of the cluster that VMware HA had a Configured Failover capacity of 1 host and the Current Failover Capacity was 0 hosts. And there was a warning stating: “Insufficient resources to satisfy HA failover level on cluster ”

After some research I came accross a nice article that explains how Vsphere calculates the weight of the virtual machines. After checking the settings of each virtual machine we discovered that one of the machines had a reservation on the memory. And one machine had a reservation on the CPU.

After correcting these settings to the defaults the errors disappeared.

So always check imported virtual machines for non-default settings.