Issues with Virtual PC 2007 and Vista

I recently upgraded my primary laptop to Vista Ultimate.  Reluctantly I gave up the security of Windows XP and the comforts of a long-lived laptop build -- not that XP was my dream OS, but I had gotten used to its quirks and I knew what to expect.  That said, it was getting long in the tooth and Vista did have a certain allure.

After the upgrade (actually a reformat and a clean install), I installed Virtual PC 2007 SP1.  When I tried to start my primary development machine, I got the following error:  "The virtual machine [VPC NAME] can't be started because it is being used by another program, or marked as read-only...."  Once my mild stoke abated, I did what every good developer does... I Googled the error and found the following articles describing other experiences with the same (or similar issue):

• http://www.aspdeveloper.net/Virtual_PC/rn-739-41912_Cant_start_any_VM_under_my_rstricted_account_(it_used_to_work).aspx
• http://doubliewkay.spaces.live.com/blog/cns!6B6E578C36BB444C!445.entry
• http://blogs.msdn.com/tolong/archive/2007/07/31/having-problem-opening-vpc-in-vista.aspx

The trouble was that all of these articles focused on security changes in Vista: either the current user (me) doesn't have access to the files or 2) for whatever reason, VPC needed to run with elevated permissions (as an Administrator).  As it turns out, I did have to adjust permissions on the files.  There were some remnants of permissions left over from my XP machine.  Unfortunately, resetting the permissions didn't fix the problem. 

In looking through various settings, I noticed a curious attribute setting - the "read only" flag appeared to be set.  Unfortuately, unchecking the read-only flag on the folder where my virtual machine was stored seemed to have no effect.  Every time I clicked O.K., closed the properties page and reopened to check the setting, I was presented with exactly the same situation -- the read only flag appeared to be set.  This lead to another Google search and another wild goose chase, that allowed me to learn the read only attribute mysteriously getting set is something that a good many other folks are struggling with when using Vista.  And, in fact, I learned that the read only attribute is only settable at the file level, not at the folder level.  You can click on the check box next to read only flag all day long and nothing will happen to the folder -- it will, however, set the flag on the underlying files.   That said, the read only flag on the individual files was unchecked in my case.

To cut the chase, I had to drop to a command prompt and use the ATTRIB command (DOS old school).  I ran the following:

ATTRIB -R G:\*.* /S 

In this case, G: is the USB drive where my virtual machine is stored.  Once I ran this command, my virtual machine started right up.  Best part... it only took five hours, lots of Google searches and a VERY frustrating, non-productive call to Microsoft support (grist for another post).

From this experience I learned a few things:

1. MSDN incidents cannot be used 24 hours a day, 7 days a week as the web site suggests (http://msdn.microsoft.com/en-us/subscriptions/bb266240.aspx).  In fact, even though development is my business and not being able to write code, for me, is by definition "being down,"  Microsoft wanted to charge me $500 to open a "critical business" incident (apparently my MSDN incident can't be used for "production" problems...curious).  If I didn't want to go that route, I had to wait until tomorrow to speak with Virtual PC support.  Why?  Apparently after being transferred to various groups no less than five times, having the CSR tell me "my machine is updating and I can't generate a case number for you" and then being hung up on, the last CSR said I missed Virtual PC support by 7 minutes -- she couldn't transfer because I'd be rerouted back to the level one CSRs.  So much for support.
2. Google is awesome for finding answers to problems -- assuming you know what to search for.  Having done this a few times, I evening got anecdotal solutions that lead to a fix.
3. Vista is generally a good operating system -- admittedly my experience beyond this problem -- has been positive.  However, these "minor" security changes seemed to have caused great consternation.

Hopefully this post will prevent others from suffering my fate.

IIR Enterprise-3 Conference

This week I get the opportunity to attend and speak at the Enterprise-3 conference, hosted by International Institute of Research and BI Research.  I'll be chairing a series of pre-conference seminars on SharePoint with Microsoft MVP Bob Mixon.  In addition, I'm presenting a session on "Sorting out the Productivity Suite Mess," which looks at the major competitors of Microsoft Office as alternatives to the market leading productivity suite.

If you're interested in broader collaboration and content management topics, check out CMS Watch Founder and fellow analyst Tony Byrne's talk "Enterprise 2.0 Technology: A Critical Evaluation."  Tony always has terrific insight into the collaboration, portal and content management space.  In addition, there will be a host of other industry leading speakers there to lend their insight into the space.

For those of you attending the conference, I look forward to meeting you there!  If you won't have the opportunity to attend, please send me an e-mail or give me a ring; I'd be happy to share what I learned as an attendee as well as sharing my presentations.

SharePoint Report 2008 Released on CMS Watch

After months of work, the new SharePoint Report 2008 has been released.  I collaborated with  CMS Watch to develop a full 190 page research report that exposes what SharePoint does well, its limitations and what businesses of all sizes need to do to be successful when implementing the tool.

I encourage you to review the sample content and review the TOC.

If you have feedback on the report, I'd love to hear it. 

Customized My Sites Exposed

We recently undertook a project to develop a global intranet for a relatively large organization.  As part of the requirements of that implementation, the client wanted to create a highly customized My Site experience.  As anyone who has delved deep into My Site customization knows, it's not straightforward, easy and may create maintenance issues long term. 

Ultimately, after working through the challenges with the client, we collectively agreed to back off the more grandiose My Site customization plans in lieu of a more simple approach.  The reasons for this change centered around the following:

• Each My Site is a site collection and dealing with several thousand My Sites when need to change elements of the implementation was not something the client wanted to deal with long term
• The original plan the client had for the My Site, while useful, ultimately removed a good deal of useful functionality provided by Microsoft
• There was some concern about "upgradeablility"

So while the we didn't create the highly customized My Site, we still implemented some limited customization that delivered a custom theme, master page and some unique navigation (to tie in the rest of the portal).  We used a technique called "feature stapling" to make that happen and ensure "compatibility" with what Microsoft would like to see in SharePoint implementations.

If you're interested in a really good set of explanations of both the My Site architecture, see a blog entry from Mark Arend.  If you are looking for the best practice approach to modifying the My Sites, see the SharePoint Product Team's blog entry.

Portals: Build it and they may come, but they won't know what to do when they arrive...

Many of our customers ask how they can spur adoption of new portal technology.  In cases where the IT department is leading the charge, we see a lot of "build it and they will come" approaches.  In much the same way Kevin Costner's character in Field of Dreams built a baseball field on his farm to encourage ghosts of baseball players to visit and play the game, IT groups hope that by simply exposing portal functionality to their end users, those same users will figure out what to do with it.  The trouble is that most employees are too caught up in their daily tasks to focus on the use of yet another tool, let alone see the opportunities for making their lives easier.

Portals technologies and other "productivity" tools come and go.  There have been lots of them over the years.  The prime example would be word processing software.  Does anyone remember the resistance to computer-based word processing over a typewriter?  Does anyone remember the fights that erupted in the executive offices when word processing centers (that existed in many large companies) were replaced by a far fewer number of individual secretaries armed with tools like the DOS-based Displaywrite word processor (an IBM created word processing packages from the early to mid-80's), WordPerfect or any of the other word processing packages of that era?  Or how about when the concept of an executive doing their own typing was introduced!  <GASP!>

The difference between a word processor and a portal is the number of functions.  More than likely, anyone familiar with a typewriter can guess how a word processing package functions.  And, with minimal fiddling, could probably figure out how to type out and print a document. 

By contrast, portal technologies encompass a number of functions -- some of them not so obvious.  The fact is and remains that IT or any organization pushing the adoption of a portal product, must help their user community.  This help comes in many forms, but here are some best practices for not only helping introduce the portal, but driving usage:

1. Pick a few Functions that have broad appeal
   Find those one, two or three functions that virtually everyone in the organization needs or wants.  The old term for this is the "killer app."  Effectively it's any feature of the portal that employees will immediately see value in using and something that you can make instantly accessible.  For example. surfacing a phone directory or portions of frequently used applications -- like electronic pay stubs.  The function needs to be something that users "need" to use as opposed to something that they could use (given the imagination to integrate it into their work).  One customer we recently spoke with implemented an employee lookup feature for a single department and it's now the most utilized feature -- by everyone from the mailroom to accounting to the receptionist.   As a result, this one feature is driving usage of the more traditional functions within their portal.
   
2. Train, Train and Train -- but do it quickly
   Most organizations miss this basic concept.  They also tend to focus on "all day" or other time consuming "big bang" sessions.  Remember that the portal is supposed to improve productivity, not create a huge hole in productivity.  By taking valuable resources away from their jobs for a day or more to try and consume loads of information, you not only create a down day, but the trainees will inevitably forget everything but the phone directory by the time they leave the session.  Instead deliver training in very small information blocks that take an hour or less.  Again, if we assume that the portal starts with a few key functions, developing training for these functions to fit within an hour or less session is not all that challenging.  If the functions are too complicated to show folks how to use them in that time, rethink the function.  Further, don't stop there.  Embed ongoing "training" into the portal, such that users can get to 30 second to 1 minute "mini-training" sessions on contextual content.  For example, if your portal has the ability to manage documents, embed links to help material literally next to buttons or links for that document management function.  A firm called Portalogiks has a product called "Virtual Training Academy" that contains 60 "modules" that fit this very description.  Their flash/shockwave-based materials are 30 seconds to 4 minute videos that present focused content around one specific function of SharePoint (e.g. how to upload a document).
   
3. Communicate Frequently
   One key to adoption is frequent and useful communication with the user base.  Users tend to forget or simply don't know that the portal may house useful functions  or, in the case of SharePoint, new approaches to accomplish the same task (see our blog entry on using SharePoint Search within Office's Research function).  Again, keep the communications focused around one topic that can easily be consumed in short order by your user community.
   
4. Flagrantly Encourage Participation
   One of my former managers, Sue Hanley, used to promote the idea of "bribery" for portal users.  Effectively, the portal won't survive without an active content contributor community and a broad consumer base.  Since the challenge is really a "chicken or the egg" problem, you should consider developing programs that encourage both the consumer and contributor communities.  In spirit of bribery, creating contests with prizes or a simple "gold" star program will help motivate both groups.  What works for your organization will probably be different from other firms, so do a little research and experimentation to figure what sells -- we've seen everything from lunch in the executive dining room to additional vacation days to iPod/Zune giveaways.
   
5. Make Incremental Improvements
   For a whole host of reasons, the big bang (rolling out all features at one time) approach simply doesn't work.  By contrast, introducing a new portal with a small set of core, broadly used functions tends to get the ball rolling.  However, without continued improvement, usage will fall quickly (or simply stagnate).  Instead, ensure that the portal continues to expand and grow.  This means not only in terms of functionality, but also approach.  There's a really good chance you won't get it right first time around.  Make sure that you become comfortable with changing how and what you're doing within the portal to keep up with feedback you're getting from your user community.  We typically recommend to clients that they make changes at least once a quarter, but no more than once a month.  This keeps any improvement scoped for a successful rollout and at a comfortable "consumption" interval, such that your users are not overwhelmed by changes.
   
6. Measure and Measure Again
   Don't guess at the portal's success, make sure you establish success criteria and gather both quantitative and qualitative data.  Use this data to continually evaluate if you're hitting your goals.  As the portal matures these measures will change, but do not become complacent.

While this list contains just six best practices, they are probably the most frequently ignored.  And, obviously, these best practices represent only a segment of what you should be doing  overall.  However, from what we've observed, just doing these six things well will vastly improve your overall success.

SharePoint Conference 2008

I just finished my time at the 2008 SharePoint conference in Seattle (admittedly a little early).  By all accounts, this conference was well attended with over 3800 participants.  Keynotes at the conference were given by Bill Gates, Kurt DelBene and three-time Tour de France winner (and first non-European winner) Greg LeMond.  In addition there were a whole slew of sessions on everything from Search Server 2008, Branding SharePoint and Microsoft's latest technology acquisition search vendor Fast.

Most significantly, Microsoft made a few announcements/general statements -- some surprising and some expected:

• "Cloud" Strategy
  Microsoft is making a commitment to introduce managed SharePoint and Exchange services to a broad range of organizations.  Up to now, only firms with 5000 or more seats could participate in their managed strategy (primarily because they were testing and hadn't worked out a multi-tenant approach).  During the conference Bill Gates and Kurt DelBene (Senior Vice President, Office Business Platform Group) announced that they are introducing multi-tenant (more than one client on a "box") managed services for both Exchange and SharePoint (in addition to rolling in LiveMeeting).  Both services will also enable single sign-on, so that historical issues with multiple authentication prompts should be a thing of the past, and a provisioning workflow for both sites and users.  Ultimately, the strategy seems to create an offering where clients can get an "enterprise" experience, where users can leverage a full compliment of traditionally on-premise applications -- like Exchange, SharePoint, Live Communications Server -- through the cloud.  Based on the demo they showed it looks promising.   You can read more here.
  
• SharePoint-like Lists in SQL
  One of the more surprising announcements by Bill Gates was a suggestion that the next version of SQL may contain a SharePoint list-like construct within a table.  Taking that a step further, it may also be possible to directly link a SharePoint list to a single table!  After speaking with some of the product team members, I think this announcement was a bit premature (they appeared to cringe when reminded of what Bill had said during his keynote), but I'll be interested to see how both SQL Server and SharePoint progress in the next release.
  
• Office Online
  While the executives stopped short of saying "we're going to give you Office over the web," they hinted we could be seeing versions of some of the Office suite products being offered over the web -- in thin-client form.  Never one to shy away from a good dig, Bill Gates took aim at Google's "consumer" centric approach, offering that they're (Google's) online productivity products aren't ready for the enterprise.
  
• Visual Studio 2008 compatible Extensions for SharePoint in June
  One source of frustration is the constant marketing of VS 2008 improvements over 2005, but a complete lack of support for SharePoint projects (with the exception of workflow).  Product team members were on hand to assure everyone that the extensions are coming, with cool enhancements over the current 1.1 release for 2005.  They estimate these extensions will be available by June of this year.
  
• Fast Search Integrated with SharePoint
  What was both disappointing and encouraging, Fast has folks onsite to discuss their work to integrate with SharePoint.  The downside was that it felt more like a marketing pitch than a "real" discussion regarding how they were going to enhance SharePoint.  That said, they did demonstrate some of the key advantages of their tool, in the SharePoint context.  In addition, they demonstrated some admittedly very slick Silverlight-enabled search result/interaction controls.  They also demonstrated a search-centric personalization approach with faceted navigation that was novel, even if it was not entirely new.
  
• Browser Compatibility and Accessibility 
  This is less of an announcement and more of what should be a red-faced admission that SharePoint isn't as cross-browser compatible as anyone would have hoped, nor as accessible.  While 2007 is vastly improved over 2003, there are still pretty significant issues on both front.  In both cases, the product team assured everyone that this was a priority for vNext (Office 14/2010ish).  As a short term fix, they're pointing people to Telerik for their "upgrades" to the rich text editor, blog editor and wiki editor.  For accessibility, there's  a "toolkit" (developed with HiSoftware), to help with the accessibility aspects.
  

Beyond the announcements, the conference was full of the usual information sessions.  Microsoft seemed to have a good mix of services partners, ISVs and  Microsoft employees (both Microsoft Consulting Services and product team folks).

I was able to attend a number of sessions, but was generally dismayed at the lack of real depth or new material.  As mentioned earlier, the Fast search sessions were only slightly better than marketing pitches for their software.  Sessions on best practices rehashed the same stuff that virtually anyone in this space would know and has already heard a dozen times. If you had attended the previous SharePoint conference, it seemed like many of those sessions were repeated at this event (more than a year later).  And, in what seemed like a bizarre event, other attendees reported that in one session, two speakers began to disagree with each other during their presentation and  the talk degraded into unintelligible babble.

There were a few bright spots however.  The sessions on the new Search Server 2008 struck a nice balance between technical detail and actual usage.  Customers like EasyJet and Chesapeake Energy did a good job of relating how their SharePoint implementations progressed and the value that they received (a similar presentation by Best Buy for Business was less interesting though).  Finally, the session on the latest version of the Community pack for Wikis was quite interesting, if only to demonstrate how SharePoint can be extended beyond it core competencies, as well as the community support behind the product.

In all, the conference wasn't a waste of time, but I would have preferred better material.  I would have liked to have seen deeper material around all of the topics (in fact for developers, the Office Developers Conference earlier this year in San Jose was the place to be).   Better titles for some session would have also helped -- a sessions titled "My Content is in SharePoint, now what" sounded like it would have been about adoption or user experience, but really ended up being about securing content.  A good topic, to be sure, but poorly "branded." Finally, the best practices session is something everyone wants (the room was packed), but they should have focused more on the specific challenges in running a SharePoint implementation like: provisioning (the explosion that can occur), the challenge around taxonomy development, when to create site definitions vs. site templates, how to train users, dealing with varying levels of Office and so on.

If you were at the conference, I'd love to hear your feedback.  Like it?  Thought it was a waste of time?  Were there particular sessions that piqued your interest?

Simple SharePoint Backup Routine

Like many of our clients, Consejo uses SharePoint to collaborate -- both with our clients and amongst our employees.  As a result, it has become a critical service to maintain -- thereby making it critical from a business continuity planning perspective as well. 

Recently a number of clients have all asked the same question:  how can we easily create backups of our SharePoint sites so that we can restore them in the event of a disaster?  While there are some good backup options built into SharePoint and some commercial vendors like AvePoint that provide backup and restore solutions, there are some relatively simply approaches that will provide baseline protection in the event of a disaster.

To begin, you should really read this post from Microsoft's SharePoint Developer Documentation Team blog on the various backup mechanisms and elements that need to be protected.  The blog entry does a great job of highlighting all of the moving parts, but doesn't provide a straightforward approach to getting the basics.  What we've tried to do, below, is provide you with an easy mechanism to automate and schedule backing up SharePoint site collections.  These backups can easily be restored in the event of a disaster.   To be clear, this approach is kind of the sledge hammer approach for site collections only and may not be right for the needs of individual sites.  However, this represents the very least you should have in place for your SharePoint environment.

NOTE:  The backup process described will backup only the site collections specified in the backup routine.  As the Microsoft article suggests, there are a lot of other parts to a SharePoint environment.   When designing your complete disaster recovery/business continuity plan, please ensure you have all of your bases covered.

Identifying the Site Collections

Depending on the type of SharePoint environment you have, your complete SharePoint environment will have one or more site collections.  When using the approach illustrated in this post, you must backup each site collection separately (or include them in a single job).  In any event, you should inventory the various site collections you want included in the backup routine.  If you're unsure what collections exist, you can look them up in the Central Administration tool by going to the Applications tab and clicking on the Site Collection List option under the SharePoint Site Management category (show in figure 1-1).

Figure 1-1 The site collection list for a given SharePoint application

If the site collections that appear aren't what you expect, make sure that you've selected the right web application, using the drop down list just above and to the right of the site collection list.

If you have found the right set of collections, take note of all the ones that are important to include in your backup routine.  You'll need the URLs to each of the collections.  Because you can have multiple collections off of the same web application, SharePoint may only show a relatively URLs; make sure that you know the full-qualified URL for this backup process.

Creating and Automating the Process

To create some basic automation and logging, we're going to create a CMD file that contains all of the commands necessary to use the STSADM command-line utility to backup one or more site collections and add some basic logging to a text file.

CMD code

The code in the CMD file should contain at least the following:

@echo off
echo --------------------------------------------------------------------- >> SharePoint_backup.log
echo SharePoint Backup Began >> SharePoint_backup.log
date /t  >> sharepoint_backup.log
time /t >> sharepoint_backup.log
"c:\program files\common files\microsoft shared\web server extensions\12\bin\stsadm" -o backup -url [URL TO YOUR SITE COLLECTION] -filename [PATH TO BACKUP FILE] -overwrite >> sharepoint_backup.log
if errorlevel 0 goto finish_backup
:error
echo An error has occured and the backup did not complete properly >> sharepoint_backup.log
goto end
:finish_backup
echo The backup completed successfully >> sharepoint_backup.log
:end
echo SharePoint Backup Finished >> sharepoint_backup.log
date /t >> sharepoint_backup.log
time /t >> sharepoint_backup.log

Keep in mind that the code contains bracketed text (like [URL]).  This bracket text should be replaced with settings specific to your environment, like the URL to your site collection or the path to your backup file.

Once you've made the modifications to your specific site, save the file to a directory on your SharePoint server; keep in mind that the log file will be written to the same directory and will collect entries over time.

Before proceeding, I would recommend testing your finalized CMD file to ensure it operates as you expect.  Once it does, proceed to the next step.

Scheduling the CMD File

There are a number of ways to schedule a process to run on a server, but Windows has a built-in scheduling system that you can use for this purpose.  The service is the Windows Scheduled Tasks.  A scheduled task can be created and run on a specific period and within the context of a given user.  To create the backup scheduled task, using the following instructions:

1. Open up the Windows Control Panel
2. Click on Scheduled Tasks
3. Click on Add New Task
4. The Scheduled Task Wizard appears
   
   
   
5. Click Next
   
   
   
6. Click on the browse button and locate the CMD file you created earlier
7. Now name the task and choose the schedule as shown below
   
   
   
   
8. Click Next
9. Enter the credentials for the user that the system will use to run the task.  In this example, we're using the Administrator account, but you should choose a user with the minimum privileges necessary to complete the backup.   Keep in mind that the user must have access to both the SharePoint site collections and the file system.
   
   
10. Once you click next, the final confirmation dialog will appear; simply click Finish to schedule the job.

To confirm that the job runs as expected, you can reopen the Scheduled Tasks option in the control panel and right click on the new job.  From the context menu, pick RUN.

After you have your job scheduled, your SharePoint site collection(s) will be backed up every time the job runs.  I would recommend at least once every 24 hours.

NOTES:

1. When I developed this routine, I made the assumption that you will get a full backup on the specified site collections every time the job runs.
   
2. The backup files are overwritten every time the job runs, so if you want historical backups, you'll need to develop some process for making a copy of the resulting backup file elsewhere.  Ideally, you could move the backup file to tape or another file system once the SharePoint backup is complete.
   
3. This backup process will backup all of the data and the security settings.  In the event of a disaster, the security context may not be the same when the site is restored.  For example, the domain used by SharePoint may change in the new environment.  As a result, you'll have to set the owner of the site through Central Administration to the new user ID who will control the site.  Then, you must grant the appropriate users in the new domain access to the SharePoint site collection.  This can get complicated if you use a lot of item or container-level security (e.g. sites within a collection or lists/libraries within a site don't use inherited permissions).
   
4. The text file that holds the log entries can fill up quickly over time.  Make sure someone is reviewing that file periodically to ensure it does get too big AND that the backups are completing successfully.

I'd love to hear your feedback on this approach and/or whether this was helpful to you.