Exchange 2013 Alerts

Posted on January 16, 2013 by Tony Redmond

I’ve been critical about the features that have been dropped from the new browser-based Exchange Administration Center (EAC) console in Exchange 2013 when compared to its Exchange Management Console (EMC) predecessor as used by Exchange 2007 and Exchange 2010.  I’ve also pointed out some areas where EAC needs refinement (or simply to have some bugs fixed) To be fair, along with its ability to support a large number of different types of device from PCs to smartphones and its coverage of on-premises, hybrid, and cloud deployments, EAC introduces some new and very useful features, among which are event notifications or alerts.

Any component is able to create an alert by writing it into the “Federated email” arbitration mailbox (the one that looks like FederatedEmail.4c1f4d8b-8179-4148-93bf-00a95fa1e042  – you can see the set of arbitration mailboxes in an organization by running the Get-Mailbox -Arbitration command).

The idea is that alerts bring issues to the attention of administrators that they might otherwise overlook and it is expected that other Exchange components will add their own alerts over time. The Mailbox Replication Service (MRS) issues alerts for mailbox migration, import, export and restore operations to EAC. The alerts notify administrators when operations start, finish, and if any problems occur. Apart from MRS, the only other Exchange component that currently signals alerts to EAC is a scan for expired certificates that is performed every 24 hours or any time that the Exchange Service host Service restarts. I expect that other Exchange components will add their own alerts over time as developers realize the value of these notifications and get the time to add alerts to their code.

Components control how they write alert information into the arbitration mailbox. Some information is written immediately an operation starts, as in the case when a migration batch or PST import begins, and is updated as the operation progress. For instance, if you start off a migration batch, MRS marks it as “InProgress”. When the batch finishes, MRS updates its notification to “Complete”. Others take a more leisurely approach because their information is not quite so time critical, which is the reason why the scan for expired certificates occurs daily.

EAC checks the arbitration mailbox for alerts that it should signal every 30 seconds. When an alert is discovered, EAC signals it to the administration. In the screen shot we can see that EAC has discovered the presence of an expired certificate. If the administrator clicks the “View Details” link, EAC brings them to the Certificates section (under Servers) to allow the administrator to deal with the problematic certificate. On the other hand, clicking the link will do nothing if you run EAC using an account that does not hold an RBAC role that allows access to certificates.

Two alerts can be seen. One for an expired certificate, the other to say that a migration batch has started

Two alerts can be seen. One for an expired certificate, the other to say that a migration batch has started

You do not have the ability to create new types of alerts, but you can fetch information about alerts or create a request for Exchange to notify specific recipients by email when a task is complete. The Get-Notification command reads information from the arbitration mailbox and provides an insight into the information available to EAC. You can compare the output shown below to the alerts displayed in the screen shot:

Get-Notification

Display Name              Type       Status      StartedBy          StartTime
------------              ----       ------      ---------          ---------
January 16 Batch (to VIP) Migration  Queued      Redmond, Tony (IT) 16/01/2013 12:23:00
Exchange 2013 Test        CertExpiry CertExpired Microsoft Outlook  14/01/2013 13:46:02

The Set-Notification command creates a request to be notified when a task is complete. For example, this command instructs Exchange that we want email sent to two users when migration batches are complete.

Set-Notification –NotificationEmails “TRedmond”, “JSmith” –ProcessType Migration

Notifications remain in the arbitration mailbox until they are removed or expire. It is the responsibility of the component that creates a notification to remove it and this is usually done when the associated task is complete. For instance, if you remove a migration batch, the notification for that batch is deleted from the arbitration mailbox. If notification items are left in the arbitration mailbox, they will be removed by the Managed Folder Assistant after 30 days because they are stamped with the AsyncOperationNotification system retention tag.

Although EAC has a bad name in some quarters, new features like alerts prove that the transition to the new console brings benefits. Once Microsoft gets around to fixing the bugs that are in EAC and adding back some of the missing functionality, EAC will be an advance over EMC. For now, well, you have to make your own mind up…

Follow Tony @12Knocksinna

Posted in Exchange 2013 | Tagged , , , , , , | 17 Comments

Exchange Unwashed Digest for December 2012

Posted on January 14, 2013 by Tony Redmond

I’m a tad remiss in posting the December 2012 digest for my “Exchange Unwashed” blog that appears on WindowsITPro.com. My only excuse is that I’ve been busy working on chapters for Exchange 2013 Inside Out (Volume 1 – Volume 2 is being written by Paul Robichaux) and should take this chance to acknowledge the help I’ve received from Jürgen Hasslauer and Sanjay Ramaswamy in sorting out some thoughts on new features that are built into Exchange 2013.

In any case, here’s the digest for December 2012:

How to Fix an Unbalanced DAG (December 27): Computer components have a habit of becoming untidy over time, or at least humans think the components are untidy whereas the components have simply behaved as they have been programmed to react to changing conditions. A Database Availability Group (DAG) is no different and databases can end by being active on non-optimum member servers. And then you run a script to rebalance matters and set your mind at ease, opening up the opportunity for yet another mince pie, or so the theory goes…

New cmdlets allow server-side control over delegate-sent messages (December 24): Proving that I was working hard on Christmas Eve, I reported on the new *-MailboxSentItemsConfiguration cmdlets that allow administrators to determine how Exchange stores messages sent by delegates. The cmdlets show up in Exchange 2010 SP2 RU4 but aren’t yet in Exchange 2013. I hope that they’ll appear there in due course.

RIM Prepares for Exchange 2013, But How Long Will BES Last? (December 20): BlackBerry Enterprise Server (BES) is everyone’s best example of a well-functioning add-on for Exchange. Or maybe not. In any case, the nice people at RIM are busily preparing to support Exchange 2013, which is nice, but the question has to be asked about how long BES can continue to last in a world where ActiveSync has become the de facto mechanism for mobile devices (and applications such as Windows 8 Mail) to access Exchange.

Exchange 2010 SP2 RU5 V2, WSUS, and WMF 3.0: quite a potential for confusion really (December 18): Microsoft has not had a tremendous record of issuing updates for Exchange in the recent past. Exchange 2010 SP2 RU5 V2 seems to be doing a reasonable job since its release, but some of the goodness was taken away by the furore around Microsoft pushing out WMF 3.0 (including PowerShell 3.0) via WSUS. PowerShell 3.0 isn’t designed to work with Exchange 2010 so that’s an unhappy combination. One wonders at times whether everyone in Redmond works with each other. Or even uses email to acquaint other parts of the company about what’s going on… But maybe they have too much email to read and process. That’s it. Kill email and the problems go away.

Outlook and Office 365: Where do SendAs items go? (December 13): This piece had been waiting to be published for a little time, but still provided some value in that it explains how messages sent by delegates are stored and how you can influence matters so that the interaction between Outlook and Exchange progresses in a satisfactory manner. The new *-MailboxSentItemsConfiguration cmdlets referred to above help even more, but only if you have Exchange 2010 SP2 RU4+ and not Exchange 2013, Exchange 2007, or even Exchange Online (yet).

Will Exchange fix ActiveSync to make sure that iOS mail can’t screw up calendars (December 11): Apple’s inability to write code using the ActiveSync (EAS) protocol to process calendar requests properly in their iOS email application has been reviewed at length in many articles. My conclusion is that Exchange should exert more control over EAS clients and stop them messing with data where they should not. Microsoft might be coming to the same conclusion, but only after several large meetings at their Redmond HQ have been screwed up because of iOS. We’ll see what happens in due course.

4 Points to Ponder About Outlook Web App 2013 Offline Access (December 6): The new version of Outlook Web App (OWA) is feature incomplete in the RTM version of Exchange 2013 because it doesn’t include functionality such as a moveable reading pane. But it does have offline access, at least it does when you have a browser that supports the feature such as the latest version of Chrome or IE10. Before you get all excited, there are some small but important details that you should consider before plunging into OWA 2013, which I attempt to explain in this article. I think I got it right. See what you think!

Exchange 2013 reaches general availability (December 4): The start of the month saw yet another announcement from Microsoft that Exchange 2013 had reached some point in its lifecycle. Following on the post-MEC announcement that the product had reached RTM, then another to say that some customers could download it, then that anyone could download a trial version from TechNet, and now general availability. Gee… but no one can deploy the blessed software because we’re still waiting for the bits that co-exist with the code that’s in product. But I guess that all the PR people are made happy by frequent good news announcements. For those that care, General Availability means that you can buy Exchange 2013 from multiple channels, including local distributors. But then ask those folk how to deploy alongside Exchange 2010 and enjoy seeing the salesperson squirm.

January is already nearly half-way through and lots more stuff has been published on Exchange Unwashed. Stay in touch!

Follow Tony @12Knocksinna

Posted in Email, Exchange, Exchange 2010, Exchange 2013, Office 365, Outlook | Tagged , , , , , , | 2 Comments

Busy week, lots of commentary about Exchange 2013

Posted on January 11, 2013 by Tony Redmond

It’s been a busy week.

First, I ran into an Exchange 2013 bug (shock, horror, cover the eyes of any watching children) when MRS stubbornly refused to export the contents of a mailbox to a PST using the New-MailboxExportRequest cmdlet. The job would get to the 10% stage where the output PST was created and MRS would start to copy items and then fail.

Some earnest dialog with some dedicated Exchange developers ensued and the upshot was that the bug was located through some testing and crash dumps. The mailbox in question was subject to a number of in-place holds, a new feature of Exchange 2013, and MRS was failing when it read a null value from some search criteria when copying data from the source mailbox. The same bug also meant that the mailbox could not be moved to another database. In any case, the bug will be fixed in due course and will appear in an update for Exchange 2013. During the debugging, I learned more about the NTSD utility than I ever wanted.

Second, it seems like a certain consensus is emerging that Exchange 2013 has some problems. MVP Michael B. Smith published his list of Exchange 2013 Gotchas to help people understand the degree of change that has occurred in the new version. Newly qualified aviator Paul Robichaux weighed in with his opinions and disagreed with Michael that Exchange 2013 is “not ready for prime time”. Of course, I have found some issues with the Exchange Administration Center that I’m not too happy with. And to wrap things up, even the normally laid-back Brian K. Winstead was moved to post that “Exchange 2013 has an image problem”.

I guess I fall between the two stools. Exchange 2013 is certainly usable and it has many good features, but feel that it is unsuitable for deployment in anything other than a brand-new green-field implementation at this point, if only because we are still waiting for Exchange 2010 SP3 and whatever version of Exchange 2007 SP3 that will be required for co-existence. Things will become much clearer when the updates for legacy versions are available and Microsoft has had a chance to address some of the fit-and-finish issues that afflict Exchange 2013 today, but I think that most will wait for Exchange 2013 SP1 before they deploy Exchange 2013 in anything more than a test implementation.

Another interesting blog post came from MVP Steve Goodman, who wrote about ActiveSync’s quarantine function and how it could be applied within companies. ActiveSync was a core part of the recent Microsoft vs. Google patent case that I was involved with in the High Court in London and during the preparations for the trial I had the chance to discuss ActiveSync with Yan Esteve, who has done an awful lot to progress ActiveSync over the last few versions of Exchange. Yan wanted me to write about how to make ActiveSync manage devices in much the same way that BlackBerry Enterprise Server (BES) allows administrators to control exactly what can connect to BES. It seems like Steve has covered the topic very well, including some script code to help populate permitted ActiveSync devices for mailboxes before controls are imposed. Recommended reading!

Elsewhere in the week it’s been a matter of pushing ahead with Exchange 2013 Inside Out. Microsoft Press dropped a bit of a bomb on Paul and myself this week by telling us that we’d need to include “alt+text” captions for each figure in the book. This information is intended to be used by readers who have limited accessibility. In essence, it’s a readable caption that informs people who can’t look over a figure about what the figure (often a screen shot) means. It’s a great idea, but it’s a lot of additional work to go back through all the chapters, locate figures, and then compose some meaningful text about each figure – in general, authors always complain when publishers ask for more work. Oh well, what must be done, must be done.

On to next week.

Follow Tony @12Knocksinna

Posted in Exchange 2013, Writing | Tagged , , , , , , , | 3 Comments

Missing functionality in Exchange 2013 EAC

Posted on January 7, 2013 by Tony Redmond

I’ve spent quite a bit of time going through the new browser-based Exchange Administration Center (EAC), the replacement for the now aged Exchange Management Console (EMC) in Exchange 2013 to create content for Microsoft Exchange 2013 Inside Out. As I’ve gone through the various sections of EAC, I’ve noted where bits of functionality appear to be missing when compared to EMC. At the time of writing, the list of missing features is as follows:

  • You can’t send a message to a user from the console. This feature depended on an email client being available on the same workstation where EMC ran.
  • The ability to join the Customer Experience Improvement Program (CEIP) is removed. You can run some EMS cmdlets instead. Most people won’t realize that CEIP signup is gone.
  • The script that counted and reported the number of Client Access Licenses that need to properly license Exchange is no more. However, the script had some numeric challenges so its absence is probably not a huge loss.
  • There is no equivalent of the Organizational Health option to provide a useful counts of objects such as servers and databases within the organization.
  • Right click to expose context-sensitive options for selected objects is unsupported.
  • The option to move the path for a database (and its transaction logs) is not available. If you want to move a database’s path, you have to do it through EMS.
  • There is no method to set logging level for a server. This option allowed an administrator to increase the amount of events logged for a particular Exchange subsystem in the application event log and was an excellent tool to use when chasing details of a problem.
  • There is no way to remove mailbox move requests through EAC. These requests have to be removed with EMS.
  • Viewing a dynamic distribution groups no longer supports a preview option to resolve the query for the group to check that the query generates the desired results.
  • EAC doesn’t have a new Sharing Policy wizard. You have to create sharing policies through EMS.
  • EAC doesn’t display details of discovery search mailboxes
  • Some of the extended calendar processing properties for room mailboxes are not exposed by EAC
  • You cannot use EAC to manipulate some of the extended mailbox properties that were accessible through EMC.
  • EAC doesn’t provide any options to manage OAB generation.
  • EAC does not capture the PowerShell commands that it executes in a log that can be later reviewed by an administrator both to learn PowerShell syntax and to understand what is actually done behind the scenes.
  • The equivalent of the EMC wizards that guide administrators through tasks in EAC do not display the PowerShell code that will be executed. This is also a loss because many administrators used the feature to copy code created by a wizard and reuse it in their own scripts.
  • Finally, many of the screens that accessed an object or a set of objects in the Exchange 2010 EMC displayed a small PowerShell icon at the bottom left-hand corner. Clicking on the icon would reveal the code that EMC would execute if the OK button was click. Again, this is a loss because it removes another opportunity for administrators to acquaint themselves with PowerShell.

It’s very difficult to replace a management console with a brand-new interface, even accepting that EAC builds on the principles established with the Exchange Control Panel (ECP) in Exchange 2010. I’m sure that Microsoft will update EAC over time to close some of the gap identified. It’s also fair to say that EAC is more functional in parts than EMC is. For example, EAC includes options to export and import mailbox data to and from PSTs.

And whatever happens, we always have EMS – anything can be done through the shell if you really try!

Follow Tony @12Knocksinna

Posted in Exchange 2013 | Tagged , , , | 15 Comments

Applying a retention policy to control Exchange 2013 Health Mailboxes (or trying to, anyway)

Posted on January 2, 2013 by Tony Redmond

Reading Jeff Guillet’s ExBPA blog the other day, I came across an entry describing a solution for the problem posed by Exchange 2013’s health mailboxes when journaling. The health mailboxes are new in Exchange 2013 and are used by the Managed Availability system to send probes to different Exchange components to ensure that they are in robust good health. Two health mailboxes are created per mailbox database so a server that supports a number of databases will generate a reasonable number of probe messages, all of which accumulate and have to be journaled, if you elect to capture messages via journaling for archive or retention purposes.

In any case, I rather liked the solution described by Jeff, which is to create a journal rule based on a dynamic distribution group with the health mailboxes being excluded from the group by filtering on a value in one of Exchange’s custom attributes. I might change the command used to create the group with the New-DynamicDistributionGroup cmdlet by including a value for –RecipientContainer to ensure that the OPATH filter picks up all mailboxes in the organization, but that’s just being picky.

The solution stops the messages created by health mailboxes from being journaled, but does nothing to stop the messages accumulating in the mailboxes. Thus, you might end up seeing something like this when running the Get-MailboxStatistics cmdlet to examine how many items are in the health mailboxes.

DisplayName                                   ItemCount

HealthMailbox79c03cca9b534959982a8773591c5f70      2416
HealthMailbox2c8bcd1322ac464d9529e82c577a876f      1251
HealthMailbox651e87104f764012aff75d0a7191464e        11
HealthMailbox79f3c6114b4a472cbd9e2cf207b179d1      2208
HealthMailboxbf5c4d3fa8a44999bf0e2e29fcb615d1      1069
HealthMailbox329cb4407fa84d01be0929879ad488d0      1349
HealthMailboxf4c968bbbbe6483faa7fc204f3634a09       752

Seeing that I am in the middle of writing the chapter on compliance for “Exchange 2013 Inside Out” The bright idea then came into my mind that it should be possible to apply a retention policy to these mailboxes to have the Managed Folder Assistant clean them out on a regular basis. In fact, a single-tag policy will do the trick, as all we really need is a default tag that removes items after a set period, which should be pretty short as there’s no good reason to retain anything in these mailboxes. I therefore created a default tag that would permanently delete items older than 2 days and added it to a new retention policy that I called “Health Mailboxes Retention Policy” (insightful naming conventions have never come easily to me).

I then applied the new retention policy to the health mailboxes by running the following command:

Get-Mailbox –Monitoring | Set-Mailbox –RetentionPolicy ‘Health Mailboxes Retention Policy’

Running Get-Mailbox with the –Monitoring switch returns a list of all of the mailboxes marked as being used by Exchange for monitoring purposes. It’s an easy way of identifying the health mailboxes within the organization. However, when I ran the command again to see whether the retention policy was now in place, no mailboxes showed up. This was very odd because the mailboxes and their underlying Active Directory user objects were still in place and were still operational as probes were still being sent and could be identified by running the Get-MessageTrackingLog cmdlet.

Health mailboxes show up where they really should not - in EAC

Health mailboxes show up where they really should not – in EAC

Even odder, I could now see the health mailboxes in the Exchange Administration Center (EAC) where they had been invisible beforehand. A little investigation revealed that Exchange now regarded the health mailboxes to be “user mailboxes”, evident by the fact that their RecipientTypeDetails properties were reported to be “UserMailbox” rather than “MonitoringMailbox”. Further probing revealed that although the health mailboxes are exposed through EAC, you can’t update their properties as the objects can only be managed by a console with a version higher than “15.0.0.0”. I had assumed that EAC was in this category because its version is 15.0.32, the RTM version for Exchange 2013, but something else is clearly getting in the way. I also haven’t figured out how to reset the RecipientTypeDetails property for the health mailboxes as the Set-Mailbox cmdlet doesn’t support this property.

I’ve reported this issue as a bug to Microsoft and no doubt will hear back from their support staff soon. In the interim, the good thing is that the retention policy is working as planned and MFA is clearing out items once they are more than 2 days old.

New health mailboxes are created normally with new mailbox databases. These mailboxes operate just like their older counterparts. However, I won’t apply the retention policy to these mailboxes until I hear back from Microsoft support! Living on the edge with the new stuff in Exchange 2013 can sometimes get on the nerves…

Follow Tony @12Knocksinna

Posted in Exchange, Exchange 2013 | Tagged , , , , , | 17 Comments

ThoughtsofanIdleMind blog stats 2012 in review

Posted on January 1, 2013 by Tony Redmond

Interestingly for me (at least), the nice people at WordPress.com sent me the following information. The most useless fact of all is that of the 397,205 views generated during the year, just 5 came from Laos. I shall clearly have to do better to encourage more Laotian readers. In any case, if you’d like to see more useless facts, be my guest and read on… All I can promise is that I’ll try and add items that provide more interesting reading during 2013.

– Tony

The WordPress.com stats helper monkeys prepared a 2012 annual report for this blog.

Here’s an excerpt:

About 55,000 tourists visit Liechtenstein every year. This blog was viewed about 400,000 times in 2012. If it were Liechtenstein, it would take about 7 years for that many people to see it. Your blog had more visits than a small country in Europe!

Click here to see the complete report.

Posted in Writing | Leave a comment

Family IT help desk calls over the holidays

Posted on December 28, 2012 by Tony Redmond

Like most IT pros, I suspect that I am the de facto CIO-and-support desk for others in the house. And so it proved over Christmas, with a variety of calls for assistance to resolve various IT issues.

First up was to get my son’s Windows RT Surface up and running. I bought this after reading of Mark Minasi’s good experience with the device as I know that Mark needs a good device on which to generate content. My son is doing a master’s degree that involves a lot of Word and PowerPoint generation. Other needs include an awful lot of PDF reading and some email. He doesn’t use Outlook so the lack of Outlook in the Student and Home version of Office 2013 packaged with the Surface didn’t cause a problem, so all-in-all the Surface RT seemed to tick a lot of boxes and would be more productive than an Apple iPad.

The Surface RT makes a nice impression out of the box. Clearly Microsoft has paid attention to the lessons of packaging excellence as practiced by its Cupertino competitor. The hardware makes a similarly good impression. It seems well designed and manufactured to a nice quality level. The only issue we had was where to locate the update for Office 2013 as the device comes with the preview edition preinstalled. Bizarrely, Windows Update seemed incapable of detecting that Office needed to be updated. At least, the super-duper Metro version of Windows Update failed in the task. The traditional Control Panel version worked and fetched the update without a bother, so that was task one of of the way as the Surface is now all set to go.

Next up was to address a strange problem with an HP dv6 PC that I had upgraded to use Windows 8. There’s always the potential for problems when you upgrade a PC to use a new version of an operating system, if only because manufacturers tend to tweak device drivers and other components to make them work well with the version of Windows that was preinstalled, in this case, Windows 7 SP1. The problem showed itself when viewing videos as the replay stuttered and definitely was much less smooth than on any other Windows 7 or Windows 8 PC that we have in the house, not to mention the MacBook Air.

Browsing HP’s web site for updated drivers for the dv6 didn’t turn up anything. The problem seemed to be with the PC’s graphics. In fact, this particular dv6 had two AMD Radeon graphics cards, with the idea being that the PC could switch between the cards depending whether it was on battery power or plugged into the mains. One of the cards (AMD Radeon 4200) was highlighted in Device Manager as having the infamous “Error 43” problem. As it turns out, there doesn’t seem to be a driver available for the 4200 for Windows 8 on AMD’s web site, so this is likely to be where the problem lay.

With no driver available, the best course of action seemed to revert to Windows 7 SP1 (64-bit Professional), which is what we did without a hitch. And lo and behold, Windows 7 cured the problem and restored video to full replay capability. This just goes to prove that sometimes a new operating system isn’t as good as an old one is, despite the best efforts of Microsoft.

Apart from dealing with new and old PC operating systems, most of the rest of the IT calls related to Apple stuff. The  directions for an Apple Magic Mouse are printed in an unbelievably small typeface, so a magnifying glass was used to figure out what had to happen (enable Bluetooth on the MacBook Air) to make everything work. Finally, an iPad had to be configured for connectivity to an Office 365 mailbox (easy because of Exchange’s Autodiscover feature).

And just when I thought the family IT infrastructure was stable,  Skype refused to connect on my own PC because Norton Internet Security took it upon itself to decide that Skype was a rogue application. Of course it wasn’t, but it is irritating when this kind of thing happens. Surely Norton knows that Microsoft spent an incredible amount of money to buy Skype and as such, it should be trusted?

This is my last post for 2012. I hope that everyone has a wonderful New Year and that you’ll come back and visit this blog in 2013.

Follow Tony @12Knocksinna

Posted in Technology | Tagged , , , , , | Leave a comment

Google loses ActiveSync patent case in London

Posted on December 21, 2012 by Tony Redmond

On December 23, 2011 Microsoft sued Motorola Mobility in the High Court in London in an action (reference HC11C04536) that eventually came before Mr Justice Arnold on December 3, 2012. I first heard about the case in May when I was contacted by Bird & Bird, a firm of London solicitors who specialize in Intellectual Property law, and asked to come over for a “chat about email”. Such an invitation was just too good to turn down, so I traveled over to London and found myself in a conference room being questioned about the history of email and my background in the subject.

Because lawyers are usually careful about revealing what they are doing and because information has to be revealed in a certain order to people who might be witnesses in a patent case, I didn’t learn what they were really concerned about for quite some time. Over the coming months the layers of the onion were gradually peeled away as we discussed issues such as the nature of email in the 1980s and 1990s, PCMAIL, and IMAP4. Of course, IMAP4 was an old friend that was well-known to me, but PCMAIL was a blast from the past. It’s referenced in the IMAP4 RFC but never saw the light of day outside MIT. Instead, PCMAIL is a description of some early work to understand the challenges of email systems that had to support multiple different clients that connect to a server.

Motorola’s case was that three Microsoft products (ActiveSync or “EAS”), Lync Server 2010, and Live Messenger infringed a patent EP 0847654 granted on 31 August 1995 that describes the process of synchronization of messages and status values between multiple pagers. Although pagers use completely different technology to email or instant messaging servers, Motorola alleged that the synchronization mechanisms used by EAS, Lync, and Messenger used the same basic method as described in their patent. For EAS, the issue was all around how email is synchronized between Exchange and mobile devices; for Lync and Messenger, it was how presence status (like “Busy” or “Do Not Disturb”) was synchronized between multiple clients connected to a single account.  Applying pager technology to modern email and instant messaging products proves that old patents can come back to life with a vengeance.

My role was to assist the court as an expert witness. Although engaged by Microsoft for this purpose, English law makes it quite explicit that an expert witness cannot be an advocate for a cause and must instead do everything possible to help the court understand the issues involved in a case. In this matter, my focus was to assess email technology that existed at the priority date (when the patent was granted) to understand whether the invention claimed in the patent existed beforehand and also whether the infringements as claimed were valid. Each side engages experts, all of whom submit reports covering topics such as “common general knowledge” (what people knew at the time) and the “skilled person” (an engineer or other working in a field) and how they would have understood the patent. These reports form evidence that is given to the judge. The experts are then cross-examined on their evidence in court, which is where the fun really starts as the debating skills of top-notch English barristers are second to none.

Google took over the case following their acquisition of Motorola Mobility on 22 May 2012. By that point the case was well advanced and it’s hard to know what advantage Google sought in pursuing it. After all, Google technical staff know a lot about ActiveSync because Google has licensed the technology from Microsoft (a point debated during trial) and as such, were probably aware of more of its internal workings than Motorola Mobility would have been (however, this did not come out during the hearing as Google’s evidence of how EAS works was very different to the way that EAS actually does). Of course, if Google won, they might have been able to get an injunction against Microsoft that prevented Microsoft distributing EAS or making it available to customers in the U.K., or perhaps forced Microsoft to pay a hefty license fee to be able to use the patent. And after winning in the U.K., they could have brought actions in other countries to achieve similar results. If you think about the number of EAS clients in use worldwide, the end result could have been very expensive for Microsoft.

From the start, my view was that the patent claims were untenable in the light of the work done in email before 1995. Synchronization was a well-known challenge as were the issues involved around working with multiple clients accessing a single mailbox. It might have been an invention to the pager community as they developed from a point where one-way pagers were the norm to dealing with two-way communications, but the invention was well known and understood in the email world. I always thought that Google was barking up a tree with no prospect of getting anywhere except having to deal with a multi-million pound legal bill. And that’s exactly what happened when Mr. Justice Arnold handed down his formal judgement this morning. All in all, the result was a slam-dunk win for Microsoft.

I really enjoyed working on this case. The legal team marshaled by Bird & Bird were easy to work with, I had a chance to debate points with some pretty sharp minds that teased out the issues to precise details, and everything revolved around the world as it was in 1995, so I could focus again on email that wasn’t so connected (the Internet was still evolving), when messages were smaller (maybe 2KB-4KB on average), and we didn’t have the volume of email to deal with that we have today. Sheer bliss!

Winning the case marks a successful end to 2012. I sure hope that next year is as good.

Follow Tony @12Knocksinna

Posted in Exchange | Tagged , , , , , , | 4 Comments

Microsoft withdraws problematic updates from WSUS, but no cause to celebrate

Posted on December 19, 2012 by Tony Redmond

Yesterday I published a note on WindowsITPro.com describing the confusion that erupted when Microsoft issued KB2506143 (for Windows 2008 R2 SP1) and KB2506146 (for Windows 2008 SP2) through WSUS, a step that normally indicates to system administrators that the fixes pointed to by the articles should be installed on servers, even if they are marked as “optional”.

As it turned out, these articles cover Windows Management Framework 3.0, which includes PowerShell 3.0. Great for Exchange 2013 servers, not so good for Exchange 2007 and Exchange 2010 servers, not to mention Small Business Server (SBS) 2008 or SBS 2011 or SharePoint 2010, which are not products qualified to work with the super-duper new version of PowerShell. The effect of installing PowerShell 3.0 onto a server is that strange things start happening, such as roll-up updates refusing to install.

It seems that Microsoft has seen the light and the two articles have been pulled from WSUS. Details are only slowly emerging, but a message posted by Doug Neal of Microsoft on marc.info says:

“As a result of these regressions and feedback from customers and experts like you, we have expired the WMF 3.0 Update for all platforms (Windows 7, Server 2008, and Server 2008 R2) as of 5:07 pm PDT.

Windows Management Framework 3.0 for Windows 7 (KB2506143)

Windows Management Framework 3.0 for Windows 7 for x64-based Systems (KB2506143)

Windows Management Framework 3.0 for Windows Server 2008 R2 for x64-based Systems (KB2506143)

Windows Management Framework 3.0 for Windows Server 2008 (KB2506146)

Windows Management Framework 3.0 for Windows Server 2008 for x64-based Systems (KB2506146)

We’re engaged in an internal post-mortem to identify and resolve the issues that led to these updates being released that resulted in the regressions.”

One hopes that the internal post-mortem asks the very salient question of how Microsoft managed to release a fix through WSUS without knowing the impact that it could have on customers, and if they did know about the potential impact, why did WSUS not come with bold, blinking, and underlined warnings? It is all very unsatisfactory and demonstrates a distressing lack of quality that has been exhibited in a number of Microsoft patches and updates in the recent past.

You’ve got to wonder why the world’s largest software company functions like this – is it a failure of management, people, or processes, or simply that Microsoft is now so large and its product portfolio so complex that it is extraordinarily difficult for any part of the company to understand exactly what happens when something is done?

The cynics among us will no doubt think that the recent quality problems are all part of a grand plot to illustrate just how much easier it is to manage IT when operations such as system maintenance are devolved to cloud-based services. It’s absolutely true that Microsoft will take care of things like applying system updates and keeping everything current for you if you subscribe to Office 365 and I assume that the folks running the Office 365 datacenters are ultra-careful in how they update their servers, but they surely depend on some of the same processes that have been proven flawed in this matter, so aren’t they also at risk?

Upwards and onwards. I wish I could be confident that another problem is not lurking in the wings. But I am not and that’s worrying.

Follow Tony @12Knocksinna

Posted in Exchange, Exchange 2010, Exchange 2013, SharePoint 2010 | Tagged , , , , , | 2 Comments

Google and Microsoft spat over ActiveSync: Hardly a war, more like a skirmish

Posted on December 17, 2012 by Tony Redmond

January 30 Update: Just on the deadline, it seems like Google has acceded to a request from Microsoft to extend its support for ActiveSync connections to Gmail until July 31, 2013. The step simply makes sense as it reduces the number of users who would otherwise be mad at Google. 

Google has licensed Exchange ActiveSync (EAS) from Microsoft since February 2009 to use as a synchronization protocol between mobile devices that support EAS. The service is branded as “Google Sync” and deploys the same EAS protocol as used by lots of mobile devices to connect to Microsoft Exchange Server. The big difference is that Gmail replaces Exchange on the back end, a situation that’s possible because essentially EAS clients and servers communicate with a series of XML requests and responses transmitted over HTTP. For those who are interested in plunging into the details, ActiveSync is documented on Microsoft’s web site, where you should probably start with MS-ASCMD, the protocol’s command definition specification (quite a mouthful).

On December 14, Google announced that it was withdrawing support of EAS for free Gmail accounts as part of an exercise curiously named “Winter Cleaning”, which sounds almost like a code name for a World War II campaign. After January 30, 2013, only Google Apps customers for Business, Education, and Government will be able to set up new devices with Google Sync. Apparently devices that are already connected with EAS will continue to work.

The announcement generated a flood of commentary, including an assertion by Paul Thurrott that this move means that “Google is declaring war on Microsoft”. I’m not so sure. Among other postings, I was amused to read Ed Bott’s comment on ZDNet.com that:

“EAS is a data exchange protocol. It’s not Microsoft code, and it has nothing to do with Exchange, the mail server program that Microsoft sells to corporations (and now to small businesses as well, via Office 365).”

Saying that EAS has nothing to do with Exchange must have come as a surprise to the ActiveSync developers that work as part of the Exchange team at Microsoft. I’m sure that the comment really meant to say that EAS stands alone as a protocol that isn’t absolutely tied to Exchange, the truth of which is seen in the fact that Google has been able to implement EAS for Gmail.

In any case, a lot of the commentary to date has been universally negative, but I think a reasonable argument can be made for Google’s decision.  Consider these arguments:

First, Google has to license EAS from Microsoft and therefore has to pay Microsoft substantial fees given the number of EAS clients that connect to Gmail. According to an article by Mary Jo Foley published at the time when Google took out its license, “the standard fee Microsoft charges its ActiveSync licensees is $100,000 “or first-year’s royalties, whichever is higher, with a per unit royalty thereafter.” Think of the number of iPhones, iPads, Windows Phones, and other devices that can connect to EAS. Now imagine how many of these are connecting to a free Gmail mailbox. And finally ask whether Google enjoys the privilege of paying Microsoft to enable these connections. I think not. On the other hand, if you’re a paying Google Apps customer, the cost of the EAS license is more than taken care of by the revenue that Google gets from Google Apps. Money talks…

Second, I suspect that the largest community of EAS clients that connect to Gmail are those using iOS devices. Google has recently launched a heap of new and updated iOS apps, including one for Gmail. It obviously makes a ton of sense to gently convince people to use the Gmail app rather than EAS because now Google controls the user interface, overall experience, and protocols. In other words, when Google decides to enhance Gmail, it can do so in the knowledge that it can issue an update to the iOS app and doesn’t have to depend on using a protocol that it does not control.

Third, being forced to use a Microsoft protocol must have been a tough pill for Google’s engineers to swallow. No software engineering company likes to be forced to buy in a component from a major competitor, especially one that forces end users to configure email clients using something labelled “Microsoft Exchange” to connect to Gmail, which is the case on iOS devices. And then there’s the small matter of having to change Gmail so that it supports the EAS protocol and the ongoing maintenance of that capability, effort that the Gmail engineers would probably prefer to expend on new Gmail features.

Fourth, making a public announcement that Gmail has dropped some support for EAS takes some of the gloss off the success that EAS has enjoyed in the mobile device market. Even though EAS has had some problems lately with iOS 6, it’s still an extremely successful protocol that is supported by a very diverse spectrum of devices including iOS, Android, Windows Phone, and Windows RT. Drawing attention to other protocols like CardDAV and CalDAV and their own iOS apps increases publicity for Google and reduces the importance of EAS.

Finally, there’s the lovely prospect of being able to tag Microsoft with the blame when an iOS client can’t configure a connection and the end user doesn’t realize why. Following the advice on an outdated web page after January 30, 2013, they select “Microsoft Exchange” and attempt to connect to m.google.com only for the connection to be declined. Because the connection type is labeled as Microsoft Exchange, the subsequent failure to connect must be Microsoft’s fault!

The change in Google’s tactics is likely to affect Windows Phone users most. If you’re in this category and have a Gmail account, you’ll be forced to use IMAP4, what is now a sadly outdated and less-functional protocol, to connect to Gmail. And if you want to synchronize with Google Calendar? Or maybe even your contacts? No problem, if you can find a Windows Phone app that speaks CalDAV or CardDAV, the protocols preferred by Google to access these data.

All-in-all, a much better solution than attempting to use Gmail with IMAP4 is to move your email account to Outlook.com and have all of your Gmail forwarded there (and the Trueswitch solution is available to move all your existing messages and other data over). Outlook.com supports EAS and Gmail forwards email superbly, so you have the benefit of continuing to be able to use EAS while also receiving email to your Gmail address. This is how I receive Gmail on my Windows Phone 7.5 device.

Make no mistake, Google and Microsoft are at war over cloud services with a huge battle raging between Google Apps and Office 365. Tempting as it is to see more than there really is in “Winter Cleaning”, I think EAS is a skirmish rather than a full-on battle. Microsoft might smart in losing some revenue from an EAS customer, for this is what Google is here, but given the importance of EAS to Microsoft’s overall mobile device strategy (for example, look at how the Windows 8 Mail App uses EAS), I rather think they will quickly get over this bump. Windows Phone users will be less sanguine, but given the numbers of these users in terms of the overall Gmail population, I doubt that Google is too worried. And anyway, as pointed out above, an obvious solution exists for Windows Phone users, so there doesn’t seem to be too much to worry about really… So why all the fuss?

Follow Tony @12Knocksinna

Posted in Cloud, Email, Exchange, Outlook, Technology | Tagged , , , , , , | 4 Comments