The haunt for knowledge

This is a follow up of the post regarding RPC errors.

First step to obtain a basic understanding of how MSRPC works can be found here. Microsoft has implemented their own version of RPC that they call MSRPC. Wikipedia puts the MSRPC implementation into a historical perspective and relates it to the source spesification called Distributed Computing Environment (DCE) from the Open Software Foundation.

I haven't yet found any specific information about how the Client- and Server Stubs are implemented in the Dynamics AX 3-tier architecture, but I'll try to dig a little bit more into this.

In addition to this, Microsoft recently also have published some KB articles on Partnersource that throws some more light on the issue I initially was blogging about. I'll try to find the URL's and update this post with the direct links when I get the time.

Update 2009-02-18:

Valuable links

Technet Magazine (article written by Zubair Alexander)
MSDN (Error Codes 1700 - 3999)
Florian's Weblog

RPC errors

One of the areas I find a little bit hard to understand, is the RPC exceptions logged in the event log (application). This is of course specific to AX 4 and AX 2009, but in my opinion Microsoft has done a pore job explaining the prerequisites and relationship between the RPC errors logged as Information and Warnings in the event log.

Sample (AX 2009 on Windows Server 2008)

I have studied this somewhat, but I have never found a good solution to generalize into best practice when installing AX. I understand that something impacts the way a client calls a Remote Procedure being executed on the AOS, but I don’t have a real understanding of the possible reasons. It could be a number of factors ranging from authentication, x++ code/logic (run at property), network packets, time outs, network drivers etc. but I really miss some basic guidelines on how to account for this. Could it also be some registry settings/tweaks needed? If yes, Microsoft should in my opinion indeed help the partners to better understand the prerequisites since this is all about making AX operating without any disturbance and “hidden” issues.

My challenge to eventual readers is to share your experience and help defining a best practice for this area. I would happily share the result of a joint community effort on this blog. If anyone remember the old “Damgaard Technet”, I really miss a common channel for addressing technical issues that I think a lot of partners around the world are experiencing, I don’t understand why every partner must invest in a lot of R&D activities to solve these kind of issues especially since it’s often about not documented prerequisites!

So please donate your experiences as comments to this blog and I promise to compile a best practice that everyone can use to better avoid RPC issues. Since a lot of unpleasant spam occurs in blogs, I have to moderate your comments before publishing but anonymous comments are allowed so keep them coming.