Things hard and not so hard.... RSS 2.0
# Saturday, January 06, 2007

I thought I'd share this festive season laugh being Christmas and all.

For the home I bought a set of Solar Powered Christmas Lights. I returned home with my purchase and instantly got huge points with all and sundry (so I might, just might be in the positives now)

I hung them up around the yard and discovered.....
the problem about these lights......think about it....solar powered lights (usually come on at night).

Let me just say that there's not one Christmas photo with these lights ON......so much for the points...easy come...easy go :-)
Saturday, January 06, 2007 11:05:38 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
Other
# Friday, January 05, 2007

A great buddy of mine gives blood, sweat and (I'm sure) tears to helping the community getting stuck into in VSTS.

Anthony has run breakfasts in Brisbane and Sydney, seminars, labs etc...all in his own time and for the love of it.

It's great to see such an expert who is passionate and has now been rewarded from Microsoft by being awarded a MVP


Anthony works with Breeze Training to offer a comprehensive VSTS Workshop. He constantly is refining/updating the workshop to reflect the latest updates and additions capable within a VSTS based solution. (I'm sure he sleeps at some point)

Well done AB!!!! Keep up the good work.
Friday, January 05, 2007 10:46:28 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
Other | Training
# Thursday, January 04, 2007

It's going around and a bit like a 'who dunnit' (sort of). In short - someone 'tags' you and you then have to reveal 5 things about yourself you normally wouldn't.

VSTS Guru Anthony Borton did me the honours and now I'm tagged. (thanks!)

So here goes:
1. I won $300 dollars when I was 15 on a radio competition for the longest toe(5.5cms) in Wollongong (where I grew up) - the comp. was on for the day and you had to go into the radio station for them to measure. My record stood from 10.30am onwards.

2. During my school holidays I used to sell Watermellons on Bondi beach in the mornings - had something going there.....

3. When I was 7 my dad dressed me up as a girl with a wig (we have a womens fashion shop in the family) etc. and took my sister and I to the local circus - as girls got in for free. (My therapist says I'm doing well now and nearly over it)

4. I'm left handed - "left handers are the only ones in their right mind" (something like that - I had a t-shirt when I was 5 with this. Yes using right handed scissors all your life is tough)

5. For my very last exam at Uni I studied for nearly 4 weeks for this last exam. Rode my push bike into the exam - panicked as the exam room was locked and no one else was there. I was 1hr early!! Got into the exam and during reading time didnt recognise one question (it was going to be tough I thought).

I discussed the exam with one of the supervisors only to be told that my lecturer in this subject stream DIDNT HAVE FINAL EXAMS!! We were assessed during the term.........I was back riding my bike again at 9.30am!

Now which island did my friends go to....

My five people I've decided to tag

  1. David McGhee - great guy now with Microsoft with a passion for just about anything.
  2. Chris Vidotto - Microsoft Technical Specialist - BizTalk. He's whooping me on xbox live at the moment.
  3. Mark Daunt - fellow BizTalk/MOSS/SQL expert - found sleeping in some bizzar places.
  4. Clayton James - Loving Sharepoint and being a MCT
  5. Mark Burch - Biztalk Support Engineer, and co-founder of the Sydney BizTalk User Group.

Looking forward in seeing what you've got folks.

Thursday, January 04, 2007 6:20:49 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [4] -
Other
# Tuesday, December 26, 2006

All full up, eaten several families out of house and home and we all slept under the tree waiting for Santa...early mornings.....:|

Best wishes and see you all in 2007 for an even better year!

Tuesday, December 26, 2006 10:32:56 AM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
BizTalk | Events | Other
# Wednesday, December 20, 2006
I'm thinking this is a bad thing if you want dynamic resolution.

I've experienced this with both the XMLReceive and a Custom Flat File Receive Pipeline (I built).

1. Both worked a treat - '5 minutes ago' as I've been working away.
2. Went into the BizTalk Admin console and 'adjusted' a property on the pipeline at the receive location ->clicked on the '...' button next to the pipeline.
3. Even if you dont change anything and hit OK...you're affected.
4. XMLReceive + FlatFile both complain about 'blank document schema' which must be specified.
(as in this case I'd changed a few things coding....it took me AGES to come back to here)
I even explicitly supplied the correct schema as a property on the FFDASM component I was using and no go!
5. Resolution: go back to the pipeline within the receive location (or send port) and select another pipeline from the list. Click OK to close the property window, then repeat and add your original pipeline but dont go into the pipeline per instance properties pages

Why does this happen??? (yeah good question :)
Basically both the XMLDASM and the FFDASM take precedence with Per instance pipeline config properties than *anything* else you provide.
e.g. Dynamically processing a flat file - check the SDK example, but here's a pseduo version.
.....
//Flatfile schema resolver - normal technique.
DocumentSchema docSpec = ......determine which deployed schema to use.

//this line basically tells the Disassemblers which schema to apply.
msg.Context.write("DocumentSpecName","<system namespace xml-norm>",docSpec.DocSpecStrongName);  
_myFFDasm.Disassemble(msg);   // line - *

These above lines will work for a year and a day (I've tested this) and in both 2004 + 2006, but NOT when you add perinstance pipeline configs - it seems all 'dynamic-ness' has gone out the window.

When per instance pipeline config is specified, this config data is provided as a Message Context Property (ReceivePipelineConfig) and XMLDasm + FFDasm only look at this for their values - painful.

So as in my case, I hadnt specified a Document Schema within the pipeline config as this pipeline was dynamic (c.f. like the xmlReceive).
I explicitly assigned a Schema to the FFDasm.DocumentSchema property and still got the same error as before.

Removing the Config Data did the trick :)


Given that this is the case - I think when you hit 'OK' to save the pipeline config data a warning/message of some description should come up as you may be wanting to modify 'one' value - 'omit xml declaration' for example. The rest of the properties are written in as blanks.

Hopefully I've saved you a bit of pain from mine.

Wednesday, December 20, 2006 11:25:41 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
BizTalk | Insights | Tips

Making some headway with this issue I'm experiencing.

I can now go into the BizTalk Application and configure most of the resources found within the Application. I've ventured into Microsoft.BizTalk.ApplicationDeployment.Engine.dll (found in the GAC) and was able to enumerate all the resources etc.

The only thing I havent been able to do is 'an update' to the resource properties (metadata).
I can re-add the resource and overwrite the existing ones, this time with the correct settings, but the problem here is that resources that have dependenies fail. e.g. schemas, maps....

For better of for worse right now I've located the two tables in the BizTalkMgmtDB and modified those directly - all looks good.

The interesting thing is I used %BTAD_InstallDir% extensively throughout the DestinationLocation.

Setting that at deploy time use:
msiexec /i <package.msi> /quiet /log <logfile> TARGETDIR=<your location>

Somewhere within the MSI Installer component within the Package - TARGETDIR=%BTAD_InstallDir%
(Even though all the documentation talks about %BTAD_InstallDir% being an environment variable)

 

Wednesday, December 20, 2006 2:37:16 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
BizTalk | Tips
I was coming across some issues with my WCF Service Clients and not shutting down properly.
They were throwing exceptions for various reasons and while trawling the ether I came across a great helper class (and this is where I saw the c# where clause) from Erwyn van der Meer

The problem centers around calling proxy.Abort(); or proxy.Close(); at different stages in the client proxies lifecycle.

Microsoft explain why we have arrived where we have on this - great candid discussion from the internal crew.

He discusses the problem and provides a great WCF client proxy helper class.

Here's a snippet from the Microsoft Discussion

Why does ClientBase Dispose need to throw on faulted state? (Or, what's the difference between close and abort?)

ICommunicationObject (from which ServiceHost, ClientBase, IChannel, IChannelFactory, and IChannelListener ultimately derive) has always had two methods for shutting down the object: (a) Close, and (b) Abort.  The semantics are that if you want to shutdown gracefully, call Close otherwise to shutdown ungracefully you call Abort. 

 

As a consequence, Close() takes a Timeout and has an async version (since it can block), and also Close() can throw Exceptions. Documented Exceptions out of Close are CommunicationException (of which CommunicationObjectFaultedException is a subclass), and TimeoutException.

 

Abort() conversely is not supposed to block (or throw any expected exceptions), and therefore doesn’t have a timeout or an async version.

 

These two concepts have held from the inception of Indigo through today. So far, so good.

 

In its original incarnation, ICommunicationObject : IDisposable.  As a marker interface, we thought it would be useful to notify users that the should eagerly release this object if possible. This is where the problems begin. 

 

Until Beta 1, we had Dispose() == Abort().  Part of the reasoning was that Dispose() should do the minimum necessary to clean up.  This was possibly our #1 complaint in Beta 1. Users would put their channel in a using() block, and any cached messages waiting to be flushed would get dropped on the floor. Transactions wouldn’t get committed, sessions would get ACKed, etc.

 

Because of this feedback, in Beta 2 we changed our behavior to have Dispose() ~= Close(). We knew that throwing causes issues (some of which are noted on this thread), so we made Dispose try to be “smart”. That is, if we were not in the Opened state, we would under the covers call Abort(). This has its own set of issues, the topmost being that you can’t reason about the system from a reliability perspective. Dispose can still throw, but it won’t _always_ notify you that something went wrong.  Ultimately we made the decision that we needed to remove IDisposable from ICommunicationObject.  After much debate, IDisposable was left on ServiceHost and ClientBase, the theory being that for many users, it’s ok if Dispose throws, they still prefer the convenience of using(), and the marker that it should be eagerly cleaned up.  You can argue (and some of us did) that we should have removed it from those two classes as well, but for good or for ill we have landed where we have. It’s an area where you will never get full agreement, so we need to espouse best practices in our SDK samples, which is the try{Close}/catch{Abort} paradigm.

 

Brian McNamara [MSFT]



Wednesday, December 20, 2006 1:23:20 AM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
.NET Developer | BizTalk
Came across some interesting code the other day and I must admit I hadn't seen the Where clause used like this before:
using System;

class MyClassy<T, U>
where T : class
where U : struct
{
}
And according to the Microsoft Definition found here

where (C# Reference) 

The where clause is used to specify constraints on the types that can be used as arguments for a type parameter defined in a generic declaration. For example, you can declare a generic class, MyGenericClass, such that the type parameter T implements the IComparable<T> interface:

public class MyGenericClass<T> where T:IComparable { }


Wednesday, December 20, 2006 1:10:17 AM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
.NET Developer
# Tuesday, December 19, 2006

I recently came across an interesting page - http://msdn.microsoft.com/library/default.asp?url=/library/en-us/BTS06CoreDocs/html/ea7038dc-4740-4c0a-b6a1-08bc22f42bc2.asp

That talks about what is allowed/not-allowed when consuming (and ultimately exposing) WebServices for use with BizTalk.

A handy reference - the interesting one is the 'Any' element not being allowed.
I suppose the alternative is to look at the BizTalk generated WSDL for WebServices that expose messages of Type XmlDocument and see what is in the actual WSDL.

Have fun - 6 more sleeps till Christmas!

Yay!

Tuesday, December 19, 2006 11:45:10 AM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
BizTalk | Insights | Tips
# Monday, December 18, 2006

Usually when referencing an external website we use http://...... and we lose all the security etc. information that comes back from our crawls. (In contrast to the File://c:/docs/... etc etc)

A handy tip you can do is IF you know the site is a sharepoint site you can use the SPS moniker so the indexing service uses the Sharepoint APIs to contact the site as opposed to the http:// protocol handler.
(1) for SPS v2 use SPS://.....
(2) for SPS v3 use SPS3://....


Enjoy.

Monday, December 18, 2006 3:39:49 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
MOSS | Tips

When I was onsite recently doing a sharepoint migration to MOSS 2007 from WSS V3 - for whatever reason all the servers (and databases) were blown away and all the client had was one content database of their site.

(Who said client sites werent exciting)

My solution here:
(1) install and configure the farm independent of the original content database
(2) Setup the the Shared Services for the Farm
(3) Extend 1 virtual server and create a new content database.
Here's the trick....
(4) From 'Application Management'->'Content Databases'....add another ContentDB
(5) Now select the original ContentDB as an additional one to add. At this point when you click 'OK' there should be two in the list
(I had to actually drop to the command line as this content database addition was going to take more time than the webinterface allowed.
stsadm -o addcontentdb -databasefile:..... -databaserver:..... -url:......)
(6) From the list of two Content Databases - take the non-original ContentDB offline and then remove it.
Viola! Worked a treak...I had to sort out a couple of things around links etc.

(I did try a few other techniques first and got a whole bunch of errors around 'object not in the correct state')

Cheers,

Mick.

Monday, December 18, 2006 3:34:43 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
Office | Tips | MOSS
# Thursday, December 14, 2006

On the whole - BTSTask.exe is a great improvement over BTSDeploy.exe in 2004.
A couple of things I wouldnt mind seeing in this tool - the ability to start/stop deployed applications.

Big Gotcha
Something I came across after my large scripting effort.....it goes something like this....
(1) 3 biztalk applications - a 'Core' and Two others.
(2) For the 'Core' apart from Schemas, Maps, Pipelines + Orchestrations; I had 1 COM Assembly, 2 custom adapters, 4 custom functoids and 3 custom pipeline components

So a reasonable sized deployment that needed to easily be deployed into test/production.

My line of thinking was - "If I could get all the associated *.dlls deployed into the BTS Core Application wthin Development...all would be good."
So as we know we can go through the BTS Admin Console and add resources/files/bindings etc. to our application (with various options), that way when we say "Export to MSI..." it's self contained.

The PROBLEM is in the 'Destination Location'....
Using the BTSTask AddResource..... setting the '-Destination:' parameter works a treat (IF your destination location exists within development environment!)

Let me ellaborate....
Development:
e:\projects\<project name>\BizTalk\
     - maps
     - schemas etc....
Associated *.dlls - found e:\projects\<project name>\CommonBin\Release

Testing/Production:
d:\Applications\<project name>\CommonBin...etc. etc.

So the drives are different and what's more, there is NO d-drive in Development....hmmm....I thought.

(I didnt have a 'demo' project to highlight this....so I've removed company info from below)

Where I want to focus is the 'Destination Location'

These 4 assemblies are deployed using VS.NET 2005 straight from the developers desktop.
(When using BTSTask AddResource....-Destination:<loc>   - loc has to exist at time of adding and 'exporting MSI' - bts validates)

Export to MSI...fine MSI finally created.

Installing the MSI file in Production/Testing
Upon performing MSIEXEC /quiet /i Core.MSI FOLDER=d:\Applications\<Project Name>\CommonBin\
I ended up getting a 'msi package deployment' - a guid as a foldername with *.CAB files underneath. No *.dlls etc to be seen.

Importing Into BizTalk
Went to BTS Admin Console and did ImportApp - all looked good.

Then went to the D-Drive and found no new files?? where were my biztalk files? gac-ed files etc?

The ones that needed Gac-ing - found copied to the GAC
The BizTalk ones Schemas, Maps etc - found in E:\projects\<project name>\<development project path>\Bin\Deployment\.....smooth! :-(

As far as I can tell this is attributed to the already existing 'destination location' within the MSI on the BizTalk artifacts.

So the reason why it's soooooo close is that if we could override this (i.e. the above FOLDER= parameter takes effect) then all would be sweet in going from environment to environment.

As it stands at the moment, I'm deploying all the files to Testing building the MSIs there and then deploying to Production with all the correct paths hopefully.

Thought I'd save you some tears.

Thursday, December 14, 2006 1:23:23 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [1] -
BizTalk | Insights | Tips
# Tuesday, December 12, 2006

Buried deep down in one of the Install Guides (multiserver, pg 23) I came across a section that shows you where to change the BAM alert email formats.

Basically there are two files - emailNotification.xslt and fileNotification.xslt both found in the ...\Microsoft BizTalk Server 2006\Tracking directory.

If you want to split the function of BAM alerting to different machines (the Provider, Generator and Distributor within the Notification Services) then there's a little step you need to do, to tell the BAM Alerting Event Provider where these new XSLT files are.

Within the Tracking Dir:

1. Run ProcessBAMNsFiles.vbs - this creates a *.adf file.
2. Edit the *.adf file to point to the new names/location of your modified *.xslt files.
3. ReRun ProcessBAMNsFiles.vbs so it picks up the modified *.adf file and makes the changes to the BizTalk Runtime.
4. Restart the BAMAlerts Windows Service.

Handy one - especially being able to modify those emails.

Tuesday, December 12, 2006 9:18:43 AM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [2] -
BizTalk | Tips
# Sunday, December 10, 2006

From students one of my most asked questions about MOSS is...."Will it handle an upgrade?"...something along the lines of .."...you know we have a medium installation of 183 distributed V2 sites with central portals, index propagation and search."

So I thought I'd outline some of my initial upgrade experiences from MOSS RTM.
(On this install - used by a small-medium sized company, they were up for technology so we had SPS2003+SP2...->MOSS Beta2 installed for a while. The RTM install wouldnt kick off until I removed Beta2)

After the product key in the setup I got these upgrade screens....(once the product is installed we can always run the psconfig.exe to start the 'wizard')

I'm doing all at once here - as this company is pretty standard in that fashion. Customisation here weigh in at about 10% of the overall Sharepoint site functionality.

 

Here's the screen for the File Locations - I moved these out to D-Drive (separate disk) and all the Indexing will be hit pretty hard.
Rule of thumb is to allow 200% space increase for the indexed content. So if you're indexing 100MB, then 200MB free for indexing is recommended.
Actually the search/index process in MOSS is pretty well defined and taking its roots from 'Indexing Service' many many moons ago. The indexes (known as Catalogs - you can access these through the APIs etc) are stored in a highly optimised format. Compressed! And the process of 'indexing' content can be 'tweaked' through things like 'the number of Word lists' in memory, Shadow Indexing and finally the master Index.

Some one was smiling....:)

 


So at this point I have all the binaries installed and all the appropriate files registered.
The Wizard will go and (hopefully) create the first Admin Site, which from there I can provision the actual Content based sites and the Shared Services (search, excel, mysite etc)

(damn...got hit with a 'we need to reboot to continue...' - so rebooting now.)

Sunday, December 10, 2006 8:37:06 PM (AUS Eastern Daylight Time, UTC+11:00)  #    Comments [0] -
Office | Tips
Archive
<January 2007>
SunMonTueWedThuFriSat
31123456
78910111213
14151617181920
21222324252627
28293031123
45678910
Blogroll
 AppFabric CAT
AppFabric Windows Server Customer Advisory Team - New Blog.
[Feed] BizTalk 2006 - Windows SharePoint Services adapter
BizTalk 2006 Sharepoint adapter!!
 Breeze SharePoint 2010 Bootcamp
Breeze SharePoint 2010 Bootcamp
[Feed] BTS 2006 R2/EDI
[Feed] Chris Vidotto (MS BTS Legend)
Needs no intro....
[Feed] Clayton James
Special Rants and a great perspective on all things....
 Mark Daunt
BTS/SPS/.NET GURU!!!
[Feed] Rahul Garg
National MS BizTalk TS - guru and great guy
About the author/Disclaimer

Disclaimer
The opinions expressed herein are my own personal opinions and do not represent my employer's view in any way.

© Copyright 2012
Breeze
Sign In
Statistics
Total Posts: 517
This Year: 10
This Month: 2
This Week: 0
Comments: 259
All Content © 2012, Breeze