Exchange 2007 IMAP RFC822.SIZE and Pine/Alpine

A common problem with Exchange 2007 IMAP when used with Pine/Alpine is alerts indicating that a message has shrunk. What is happening here is that Pine is retrieving the RFC822.size from Exchange, downloading the message and comparing the real message size with the size Exchange 2007 provides. Sadly there often is a huge discrepency, sometimes by as much as a factor of two. Interestingly Pine only checks for size decreases in case of data loss during copy hence you only see this message appear on some emails, not all. If Pine checked for message decreases and increases you would see it for pretty much all emails.

Microsoft have so far explained that the size returned by Exchange prior to Exchange 2003 was based on the MIME message stored in the STM database. As the STM database no longer exists in Exchange 2007 the RFC822.size is calculated based on the MIME skeleton and the original MIME message size. This is done for performance reasons. I have asked Microsoft for clarification as to why the original MIME size of the message is vastly different to the regenerated MIME message and am waiting for clarification.

There is an Alpine code change that will supress this message (if you are happy to lose the safety check), or you can apply a change to Exchange on either a server or on individual mailboxes. The setting is nothing secret but documentation is very scarce. It is worth mentioning that making this change has a performance penalty as it forces Exchnage to regenerate each message as MIME and then send the size of the regenerated MIME message every time a message size is read.

For the server:

Set-ImapSettings -EnableExactRFC822Size:$true

For each user:

Set-CASMailbox "username" -ImapUseProtocolDefaults:$false -ImapEnableExactRFC822Size:$true


Although this solves this particular problem, sadly other issues persist with Pine and Exchange 2007 like Pine crashing when opening messages with Digital signatures and Pine not handling multipart messages correctly. It seems that although these issues only appear in Pine/Alpine (and perhaps Mutt), it appears to be Exchange that is failing to follow the RFC standards (and Pine is a stickler for standards!).

I'm sure we will get there soon!

Exchange 2007 SP1 RU5 - IMAP Error 4999

Hi,

Like many other people we have been suffering from our Exchange 2007 SP1 RU5 IMAP service constantly falling over with the following error:

Watson report about to be sent to dw20.exe for process id: 6848, with parameters: E12, c-RTL-AMD64, 08.01.0336.000, M.E.Imap4, M.E.D.Common, M.E.D.G.OutboundCodePageDetector.AddText, System.ArgumentNullException, 3e79, 08.01.0336.000. ErrorReportingEnabled: False

The good news is that we have put a stop to this by isolating it to a specific user.

There are many records online about this being a known issue, which it seems is indeed the case and we have received the fix from Microsoft (which we intend to try on the 27/02) - KB960292. I will post how this goes. The problem relates to a null ID being presented during login that the IMAP service should ignore, but it fails instead.

Anyway, what I wanted to share is that we have isolated this issue to the Sea Monkey 1.1.14 client. All three installations we know of are causing this issue and we have worked around it temporarily by preventing the users from logging in using IMAP (they are using OWA temporarily instead).

Good luck if you are experiencing this problem.

Rob

Increasing Redirect to Voicemail Timeout on O2

When I received by lovely new HTC Touch HD I found I kept missing calls as there was a slight delay between the caller calling and the phone starting to make a sound. I contacted O2 and they told me to run the following command :

**61*901*10*??#

?? is the delay you want in increments of 5 up to 30. I went for:

**61*901*10*25#

Works a treat.

I hope this saves someone a bit of time.

This is the home email address of this recipient. It cannot be removed

It has been a while since I posted, but I came across a nuisance error this morning: "This is the home email address of this recipient. It cannot be removed". The address that could not be removed was not the primary address and what's more this was a mailbox, not a contact, so I could not just delete it without disrupting the user. I needed to remove the rogue address as it had been created in error and was a domain that was not managed by Exchange. Anyway, I launched ADSIEdit.msc, found the attribute TargetAddress (http://msdn.microsoft.com/en-us/library/aa487600(EXCHG.65).aspx) and cleared it. I could then go back to ADUC and remove the rogue address. After about 20 minutes mail started to be delivered to the mailbox once again.

Using the Safari Browser with the Citrix Web Interface 4.0

Citrix provide a very helpful article at http://support.citrix.com/article/CTX104367 describing how to customise the Citrix Web Interface 4.0. to support the Mac client on Safari. This works well, but it is not the whole job as when the user logs off and back on again, they will find their client is set back to Java. It also does not present the user client installation prompts in the message centre until they are logged on.

It is fortunately very simple to complete the job. Follow the steps in the Citrix article and then do exactly the same for the following files in the auth folder as you did for the same named files in the site folder: include.cs, embed.cs, install.cs.

Works a treat. Obviously all changes should be tested in a dev environment first and don't forget to read Citrix's disclaimer on the webpage linked above.

Windows Live Mesh

I recieved my invite to Live Mesh today and found it would not install and reported that I should check my network settings: "Installer encountered errors: Please verify your network settings and attempt to reinstall".

Well all was fine with my network settings. I tried setting my locale to US as suggested on some site and ensured that UAC was switched on. I then tried changing my auto proxy configuration script to a hard coded proxy (in IE connection settings) and Live Mesh then installed. So if anyone else sees this error and you are using an automatic configuration script for your proxy, that is why.

Hopefully Microsoft will fix this at some point.

SPSS and Appsense Environment Manager

I am having trouble with SPSS 15 running on Citrix Presentation servers which also has Appsense Environment Manager running. Basically, Environment Manager kills the SPSS processor. I can reproduce this consistently and fix the problem by either disengaging the EM process, removing the EM DLLs or simply removing any applied EM policies.

I have had a call open with Appsense about this for a few weeks. The only suggestion they have made so far is to upgrade to the most recent version. I did this (7.2.227), but it made no difference.

Currently I am trying SPSS 16.0, but not having much luck as our license for SPSS 16.0 is not authorised for use on a terminal server......... Hopefully I will have a call from SPSS soon to establish how I can run this test.

Has anyone seen this issue?

I will keep this post updated, but it looks like we are going to have to switch of Environment Manager on our Citrix servers and live without the good things that it does.