by Michael B. Smith, MCSE/Exchange MVP

It’s been a tough couple of weeks in the Exchange Server world.

First, Rollup 6 for Exchange Server 2007 RTM (release-to-manufacturing, that is – the original version with no service packs) was released before Rollup 1 for Exchange Server 2007 Service Pack 1. On the ‘nets, a lot of people complained that they felt like they were being penalized for having installed service pack 1! This may just be an indication that some people like patching too much! :-)

Rollup 6 for RTM was released on Monday, February 25th. Rollup 1 for SP1 was released on Thursday, February 28th. Just a three day difference, but for most larger companies it was probably the difference between a weekend maintenance window for the February 29th weekend versus a weekend maintenance window for the March 7th weekend.

However, that rolls us (pun intended) straight into the next problem, which undoubtedly had many people scratching their heads if they had installed either rollup by February 29th…. Exchange Server 2007 (all releases) had a leap-year bug! If the System Attendant service was restarted anytime on February 29th UTC (UTC is Universal Coordinated Time – it used to be called GMT for Greenwich Mean Time) then everything that would cause an address list to be updated (i.e., new-mailbox, disable-mailbox, setup.exe) would generate an error: “The Exchange server address list service failed to respond. This could be because of an address list or email address policy configuration error."

You just have to wonder, after last year’s fiasco with the DST changes, how this problem made it through Quality Assurance. The Exchange Team assures us that the problem is fully understood and they also say (one must presume tongue-in-cheek) that they will have a repair available before the next time this problem occurs in 2012.

Our next issue was caused by MS08-003 – a security patch designed to protect domain controllers from a denial of service problem that could occur when certain badly formed requests for LDAP searches were submitted. This patch was released on February 12th, but it is a testament to the fact that it takes most companies several weeks before patches get installed, that it didn’t really begin to show up until the first week of March. In “larger” environments – those where there are more than about 1,000 users, Exchange Management Console or Exchange Management Shell could throw the following error: Active Directory operation failed on <domain controller name>. This error could have been caused by user input or by the Active Directory server being unavailable. Please retry at a later time. Additional information: Additional information: The directory service encountered an unknown failure. Active directory response: 000020EF: SvcErr: DSID-020A0F27, problem 5005 (UNABLE_TO_PROCEED).
This error would occur when either tool was attempting to display more than 1,000 items. A patch is expected “soon”.

Finally, in March 5th’s msexchangeteam.com blog posting, Microsoft made a statement that has been true for quite some time, but caught quite a few people by surprise. When running Exchange Server 2007 service pack 1 on Windows Server 2008 (which is something very few people are doing outside of Microsoft so far…) there is no in-built mechanism for taking online Exchange database dumps. NTbackup is not part of Windows Server 2008, and its replacement does not support Exchange streaming backups, or VSS-backups of Exchange. A third party utility is a requirement. All I can say is: ouch!

There certainly have been some good things too! Microsoft Online Services was announced on March 3rd (that will be a topic for a future article) and on March 6th, Apple announced that they will support Exchange ActiveSync on the iPhone – to fully bring support for it into the enterprise.

I surely hope that this month is quieter than last month on critical support issues in Exchange Server!

Some bug complaints seem to run in spurts. One week everyone has a problem with the same bug, and then no one complains about it for weeks or months, until several people notice it again. The “Missing All Day Events” bug is one such bug.

This bug is caused by corrupt view. Typically, when the view becomes corrupt, the all day events are visible when Outlook is maximized but when it’s near full screen the all day events aren't visible. If you switch to a lower resolution, the events are visible. Adjusting the size of Outlook's window, or the calendar pane by adjusting the size of the navigation pane or to-do bar will show or hide the all day events.

Resetting the view will fix it, either by resetting the Day/Week/Month view or opening Outlook using the /cleanviews switch. I know from experience that the problem will recur when just the Day/Week/Month view is reset. Using /cleanviews tends to do a better job at fixing corrupt views and problems are less likely to reoccur, but it does wipe out all existing custom views. Since I have many custom views and want to avoid recreating them, I’ll tolerate resetting the view every 2 or 3 months as needed.

Disappearing All Day Events
http://www.slipstick.com/problems/disappearingevents/disappearingevents.htm
(Includes screenshots and a video of the bug in action)

Ok, so it's not really a bug but a limitation of the dual time zone display when daylight time goes into effect, and to a lesser degree, when we revert to standard time. Apparently, it’s not a very common problem since very few people complain about it, but it’s interesting. At least I think it is.

When you use dual time zones in Outlook and the default zone is west of the secondary zone, an extra hour is added to the time offset for the Sunday DST goes into effect. If you view multiple dates and begin with that Sunday, the (incorrect) offset is used so it appears Outlook has the times wrong.

For example, if your default time zone is Central and Eastern is your secondary time, when Sunday rolls around, the time scale incorrectly displays a 2 hr time difference all day. If you use a multi-day view with Sunday as the first day, the Sunday time scale is used, so it appears there is a 2 hour time difference each day.

This view anomaly occurs because the default is a zone west of the secondary and Outlook takes into consideration that DST goes into effect in the earlier time zone and adjusts the times using the wrong offset. The entire 24 hour period uses the same offset and the first day in the view determines the offset. This issue also affects the view when the time “falls back” to standard time, but only if the dual time zones are 3 or more hours apart.

Fortunately Outlook uses GMT and time zone offsets to calculate appointment times and reminders are based off the default time zone, not the secondary zone, so it’s unlikely to cause missed appointments.

“Time heals all wounds” and it will fix this display problem – until then, use the work week view if you use the dual time zone feature and the default zone is west of the secondary zone.

Exchange server's Remote Domains allow companies to control message format options and whether or not receipts, automatic replies and forwards, or out-of-office messages are sent to specific domains.

In Exchange 2007: Open the management console and browse to Organization Configuration, Hub Transport. Select the Remote domains tab. The Default domain settings apply to all messages sent and many administrator block out-of-office replies to all Internet addresses. If there is a specific domain that should receive receipts or replies create a new remote domain with the desired options enabled.

In Exchange 2003 and 2000, look in the ESM under your Exchange organization, Global Settings. Right click on Internet Message Formats and choose New, Domain to add custom settings for a specific domain.

Good news for Google Calendar users – it supports 2-way sync with your outlook calendar (or one way sync, either to or from Google’s calendar is an option). You’ll need to download and install a sync tool from the Google Calendar site –log into your account and look for the link.

Syncing is done every 10 minutes and seems to work so far. I’m testing it on an Outlook 2003 system and occasionally receive an error message that I need Outlook 2003 or 2007, but syncing works. The only other problem so far is that some recurring appointments in Outlook do not sync up to Google’s calendar.

Do you get annoyed because your new appointments default to All Day events? This is caused by double clicking on a date in the Monthly view -- appointments created this way will always be events, not timed appointments. Use the New button instead (or Ctrl+N, or right click, New Appointment) to open a new appointment. While you can select a date on the calendar before clicking New, the time defaults to the start of your business day and a 30 minute time scale, regardless of your time scale in the day or work week view.

If you use the Day or Work week view, double clicking on a time period opens the appointment form with the time and date selected. Drag the cursor across several time periods then select New appointment from the right click context menu to create an appointment covering that time period.