An interesting collision of Outlook views that I’ve lately come across -
sporadically I’ve begun receiving complaints that migrated users (both Exchange 2003 and 2007 to Exchange 2010) are having problems with default view of different shared folders (e.g Contacts or Calendar).

The problem manifested itself in the form of disappearing default view that was replaced by generic Outlook Today view.
The user who shared folders was not able to see such a behavior and only delegates were able to replicate it.

A couple of similar cases ended up with mailboxes recreation that I considered too harsh for such a problem and decided to take a deeper look into it.

The problems outlined in the incidents which ended up with mailbox recreation were lacking a few key troubleshooting techniques that I’d like to engage.
Mailbox recreation apparently is way too brutal and I personally would use it as the last resort only.

This is pretty crucial for successful troubleshooting to figure if this is the global issue for all delegates set for a specific mailbox (I wouldn’t hesitate to add a test one in case there’s only one existing originally) or the problem of view rendering for a specific delegate.

I believe this is a logical must to separate the problem into 2 possible clauses:

1. Shared folder problem
2. Delegate’s problem

I decided to go this way as I’m not that kind of mailbox recreation junkie.

It turned out the problem was not reproducible on other delegates’ accounts and was absolutely isolated to a specific person trying to open a shared folder. Apparently with this in mind you’ll never say ‘go’ to mailbox recreation.

You may find some solutions on the Internet involving running /resetfolders switch that will never help in such issue or rebuilding user account (like this link from Experts Exchange) -

“I resolved the problem for my situation by rebuilding the user account. First I exported the entire mailbox to a .pst then from Exchange
Management Console I removed the mailbox and associated user account. I recreated the account with the same name and after granting access
to the Contacts folder everything works as it should. Note: the profile on the user’s computer will detect a new user account so you will
need to copy user info into the new profile. ”

The right way is to use /cleanviews switch that does its job just fine. This needs to be run against the profile setup with cache mode off. There is a way to take care of views problems with mfcmapi but this requires some more experience with it in order not to destruct the mailbox occasionally. For end users running the switch is absolute no brainer (compared to mfcmapi tricks:)).

As a resume, expediting a bit more on possible way of resolution may help us avoid mailbox recreation unless it’s absolutely necessary. Scalpel is way better than the axe, huh?

UPDATE:

Just had the same issue on other mailbox and the steps outlined above were of no help. The problem persisted even after mailbox move and running the appropriate switches.

Given the repeated nature of the problem the case with Microsoft was opened and here are the brief details -

First off, the recommendation was to run the following cmdlet against the database where the delegate was resided:

Isinteg –s “server name” –fix –test alltests

Secondly if we still have the issue then we have to export the delegate’s mailbox data to a pst file and delete the mailbox. Recreate the mailbox and import the data back.

Since the first step required bringing the entire database online that was not acceptable at that time we proceeded with the second step that did the needful and the problem was resolved.

 

Recently I came across one of the weirdest mailbox corruptions I’ve ever seen.

Initially the problem arose as the customer’s inability to login to Outlook Web Access and later on expanded to problems using Outlook mail client while deleting the content.

OWA login page was producing nice call stack errors like these -

Exception

Exception type: Microsoft.Exchange.Data.Storage.StorageTransientException

Exception message: Cannot set search criteria in SearchFolder.

Call stack

Microsoft.Exchange.Data.Storage.CoreFolder.SetSearchCriteria(SearchFolderCriteria searchFolderCriteria, SetSearchCriteriaFlags setSearchCriteriaFlags)

Microsoft.Exchange.Data.Storage.SearchFolder.ApplyContinuousSearch(SearchFolderCriteria searchFolderCriteria)

Microsoft.Exchange.Data.Storage.RemindersSearchFolderValidation.VerifyAndFixRemindersSearchFolder(DefaultFolderContext context, SearchFolder reminders)

Microsoft.Exchange.Data.Storage.DefaultFolderValidator.Validate(DefaultFolderContext context, StoreObjectId folderId, Dictionary`2 folderDataDictionary)

And like this -

Inner Exception

Exception type: Microsoft.Mapi.MapiExceptionNotEnoughMemory

Exception message: MapiExceptionNotEnoughMemory: Unable to SetSearchCriteria. (hr=0x8007000e, ec=1008) Diagnostic context: Lid: 18969 EcDoRpcExt2 called [length=293] Lid: 27161 EcDoRpcExt2 returned [ec=0x0][length=572][latency=15] Lid: 23226 — ROP Parse Start — Lid: 27962 ROP: ropSetSearchCriteria [48] Lid: 17082 ROP Error: 0x3F0 Lid: 27745 Lid: 21921 StoreEc: 0x3F0 Lid: 27962 ROP:

Having setup Outlook profile I was able to login to mailbox and here’s the weirdo I met -

This marvelous pattern might have apparently been found everywhere (e.g in Deleted items, not only in Inbox as here). This is iteration that cannot be handled by Exchange. Moreover, this pattern cannot be deleted at all!

Neither by Outlook/OWA means nor MFCMapi. The latter throws exception too.
Here’s the solution:

Export mailbox content skipping the folder containing this weirdo.

Mighty PowerShell does it as easy as possible, see generic example below:
Export-Mailbox -TargetMailbox -TargetFolder -ExcludeFolders “\Junk E-Mail”,”\Contacts”

Re-create a mailbox, import the data.
That’s it.

 

In my Exchange Admin practices I frequently come across peculiar problems. Probably this is only of the excitements the work can give:)
Back to the interesting problem I encountered -
it was that OAB did not get updated to reflect recent changes made on user’s end. Not a big deal you say as OAB problems are not rare indeed. The blog of OAB guru may prove this standpoint easily.
Manual update on the server completed successfully but it did not help when Outlook was engaged into play – still no changes visible. Sporadically I received infamous 4010F error, strangely it hit me not on every workstation I used to test.
Given that, the app log was of an essence. And it revealed a portion of helpful data -

Event Source: MSExchangeSA
Event Category: OAL Generator
Event ID: 9334
Date:  X/XX/2010
Time:  10:30:49 AM
User:  N/A
Computer: COMPUTER
Description:
OALGen encountered error ffffffff while initializing the offline address list generation process. No offline address lists have been
generated. Please check the event log for more information.
- /o=XXX/cn=addrlists/cn=oabs/cn=user

This fffff error is frequently caused by some misconfiguration regarding OAB regeneration on the server. A little bit of research helped diagnose a problematic system path written in the registry.

In the particular case the path to OabDropFolderLocation was incorrect -
HKLM\SYSTEM\CurrentControlSet\Services\MSExchangeSA\Parameters\COMPUTER\OabDropFolderLocation pointed to a non-existent hard drive. It must point to a valid location like <disk:>\exchsrvr where <disk:> reflects real physical location of this folder I changed it to the one that was correct for given server, regenerated OAB and voila – the issue was gone.

And yes, before making a registry change like that please be extra careful as any registry modifications may be a lethal blow for your server if performed without proper care.

 

Today is a BlackBerry day for me (in other words not the best one!).
One of the most frequent woes the BlackBerry handheld activated for Enterprise may present is an inability to send out messages.The crucial step to the fastest resolution is to get the error message. To do this, simply open up the message that failed to send and scroll up and there will be an Error.

Among common issues are:

- Unlisted Message error. This error message means the Message Agent lost contact with the Domain Controller/Global Catalog Server. This kind of problem requires server access to get it fixed. In 99 out of 100 cases it will be Dispatcher service that is to be restarted. The most recent updates for BES both 4 and 5 branches should resolve these issues.
- No Service Configured. This means there is no service book to send the message. Try resending service books or re-activating the device.
- Transaction error – decryption error. Regenerate encryption keys.

This is my today’s BlackBerry FAQ:)

 

I must admit – I do not like BlackBerry. Probably this is the soul’s cry of a tech guy who repeatedly encounters different sort of issues with this technology.
As the Exchange pro I would speak about BES exclusively leaving behind BIS.
In the particular post I would like to discuss one of the possible activation failures that can well be treated on user’s side rather than on the server’s (what a rare occurrence this is!)
Okay, here’s the plot: Enterprise activation starts and never completes. The activation process may get stuck somewhere in middle of the road, say, 30% or 50% or 80% or whatever without any reasonable explanation why it gets stuck.

If you’re BES admin then this is easier for you and you may check the logs trying to find something like this:

Event Type: Warning Event Source: BlackBerry Messaging Agent BES SERVER Agent 3 Event Category: None Event ID: 20530 Date: 9/8/2010 Time: 9:05:22 AM User: N/A Computer: COMPUTER Description: {unlucky_user@badluck.com} ConstructPIMFolder – Failed to open the memo folder for user (0×80040107)

There is a good article providing the problem description.

So I followed the article and checked PR_IPM_CONTACT_FOLDER_ ENTRYID and PR_IPM_NOTE_FOLDER_ ENTRYID both in Root Container and in Top of Information Store with MFCMapi editor tool and found discrepancies in Notes:

Notes in Root container
00000000F9D0489742D7B848B39AF20360B824DE01006353D4DC26523B4587F5254CC0DD8951000017D49EB50000

Notes in Top of Information Store
0000000091120EDDF117F043AC860848CE1C39020100ACC35E4A748E594C86E473EDA151DAF30000062F03620000

Usage of MFCMAPi can be lethal for a mailbox when done by inexperienced end user so to make things easier there’s a workaround to be performed on end users’ end:

This is to run Outlook with /resetfolders parameter and it should do the trick (Start > Run > Outlook.exe /resetfolders). Please note this switch needs to be run against Outlook profile setup without Cache mode.
After that the same values should become identical.

Once the properties mismatch is fixed the reactivation should go smooth.

© 2011 EHLO Me! Suffusion theme by Sayontan Sinha