Thursday, October 27, 2016

Integration Designers and Developers Need Business Context

By Steve Endow

At the GPUG Summit 2016 conference in Tampa 2 weeks ago, I gave a presentation on Dynamics GP system integration design with Tim Wappat.

This is one of my final slides in the presentation:

Out of context, this slide may not seem as meaningful as it is in the presentation, but the idea is that when system integrations are not simplified or designed strategically, they are just glorified data imports that do not help the business strategically.

The next slide offers some goals:

Part of making system integrations successful and more valuable is designing them to meet business requirements.  It isn't just about pushing data into GP.  It's about ensuring that the entire purchasing process, for example, from requisition to PO to receipt to invoice, goes smoothly and with as little manual effort as possible.

To do that properly, the entire team needs to be aware of the context of the system integration and understand the business use cases and requirements.

Friday, October 21, 2016

Two lesser known SQL Server performance tips: Appropriate for Dynamics GP?

By Steve Endow

At the amazing GPUG Summit 2016 mega conference in gorgeous Tampa, his eminence John Lowther, MVP, hosted a session "SQL Administration Made Easy", where he discussed several great tips for simplifying SQL Server administration for Dynamics GP.

Sunny Tampa, October 2016

In his role at nJevity, John and his colleagues have to manage hundreds of SQL Server instances, so they've worked hard to optimize their SQL configurations to work well for Dynamics GP.

As a result of that session, I've kept my eye out for SQL optimization tips, and I have come across two lesser known configuration settings related to SQL Server.

Tuesday, October 18, 2016

Conversations from Summit: What do you call your partner/ISV/VAR/consultant?

As promised, I am continuing to blog about some of the great conversations I had last week at Summit in Tampa, Florida.  This one comes courtesy of multiple discussions over days, with clients as well as with coworkers and a few long time consulting friends.  But, first, a little background on me.  I did indeed become a consultant out of mere happenstance. I needed a job, I knew a little (very little indeed) about software, and took the leap.  Over the years, however, being a consultant has become part of who I am and I find it hard to imagine a different career. I love what I do for two primary reasons (among many lesser ones):

  • I enjoy helping clients get the most out of their systems and processes, becoming a "trusted advisor" that they seek out for advice, guidance, and perspective
  • I like building partnerships with clients, where I learn from them (about requirements, about business, about what keeps them up at night) and I (and my fabulous teammates of course) can impart some of my knowledge to them

The discussions I had last week at Summit circled around the client relationships I enjoy most.  And we kept coming back to the term "partner".  I love it.  It is my preferred term.  Because it suggests, obviously, a partnership of give and take and mutual benefit.  Not simply a business exchange of services for money (which is why terms like "vendor" or "contractor" don't appeal to me).  Of course, I am not talking about a word game but more the nature of the relationship between consultant and client.

Of course we have to start with a solid base of product knowledge and experience, but the truly great client/consultant partnerships extend in to so much more.  Where we as consultants can benefit clients in ways that extend beyond how to configure a module or cut an AP check.  And on the client side, they can contribute greatly to our understanding of their business, industry, and needs...which makes it that much easy (and dynamic) to serve them well.

So I challenge you to look at your relationship with clients or with consultants (depending on your role), and see how you can contribute to making it a partnership. One way to look at it is to make note of how you refer to them (either client or consultant), is it in a defensive/combative way (preparing for an issue)? Is it distanced, lacking in trust? And then commit to moving towards a partnership-- sometimes that means a "clearing the air" conversation, or discussing specific expectations from both sides to improve the relationship, or simply reaching out regularly with questions that extend beyond error messages.  On both sides, it requires a fair amount of good faith, trust, and expertise. Or at least it does in my opinion, do you agree or disagree?

Christina Phillips is a Microsoft Certified Trainer and Dynamics GP Certified Professional. She is a director with BKD Technologies, providing training, support, and project management services to new and existing Microsoft Dynamics customers. This blog represents her views only, not those of her employer.

.NET Error: Side-By-Side Configuration is Incorrect

By Steve Endow

A client recently installed a new .NET eConnect integration for Dynamics GP.  After configuring the integration, they received this error when they ran the EXE file:

They also saw this message in Event Viewer, which helped narrow down the cause.

I quickly reviewed the exe.config file, but couldn't find any obvious errors.  I tried opening the XML file in Internet Explorer as a quick test, and did see that it didn't load properly, so there did appear to be an issue with the file format.

I then opened the config file in UltraEdit and used it's XML validation, but while it did say there was an error, it gave me a misleading description of the error, pointing to a node that was valid.

So I finally resorted to opening both the customer's config file and my dev config file and flipped between the two to compare line by line.  I finally found the problem.

One of the closing value tags had been accidentally deleted when the file was edited.
Once I added back the close value (< / value >) tag, the import worked properly.

You can also find him on Google+ and Twitter

Friday, October 14, 2016

Conversations from Summit- Security and Asking Why?

Sitting on a plane flying home from Tampa is a great time to reflect. I have always enjoyed Summit for a variety of reasons beyond the sessions like...
  1. Reconnecting with long-time consulting friends, vendors I have known longer than I have been married, former students, and clients
  2. Inspired  conversations about what we do, how we do it better, and why we are so collectively passionate about how software can help businesses grow and individuals prosper
  3. I always leave having learned something, some years it is functional/technical know-how and other years it might be a perspective change or a piece of software that solves a common issue
One of these inspired conversations happened this week when I was sitting with a group of MVPs and others in the Mekorma HUB discussing who is trusted in an organization.  So I wanted to share a few of those points, and expand the question a bit.

It is not uncommon for us (as consultants) to get requests to restrict security for users.  Totally common.  What gives us pause from time to time is the role of some of the individuals that are being restricted like...

  • Payroll processors
  • Business analysts
  • HR managers
  • Financial managers
It is not so unusual to have security in place to protect processes and demonstrate the process control that auditors like to see in organizations.  But sometimes it is the specifics of what is restricted that makes me think.

For example...
  • With payroll processors- restrict their ability to see any dollar values
  • For business analysts- restrict their ability to see certain financial numbers, or detail information behind them
  • For HR managers- restrict their ability to see payroll information
  • For Financial managers- restrict their ability to use reporting tools, like Smartlist
I don't mean to suggest that there are not valid reasons to do all of the above.  But I do think that if you find yourself doing (or requesting these things), you should pause for a moment to ask "why" and do a little due diligence on the goal.  Often times, when we have these "pause" conversations, we find that the reason falls in to one of two primary categories: trust or training.  Here are some of the specific reasons we hear, categorized accordingly..

  • Trust
    • Concerns that employee will not be discrete/act professionally with information
    • Executive level edicts that certain data be restricted to limited individuals
    • Internal behavior that suggests employee would be upset with what they learn
    • Employee is new to organization
  • Training
    • Concerns that the reporting that will be created will be incomplete/invalid
    • Experiences where individuals "messed up" the system inadvertently
    • Wanting to avoid unnecessary questions or "busy-body" behavior
In the cases above, I encourage organizations to address  the root cause.  Limiting access doesn't address these gaps or behaviors, it just eliminates the opportunity.  If we are in the business of growing employees (which I think all organizations are), we need to have these coaching conversations, provide the training and guidance they need, to be better professionals (which in turn benefits the business in not only outcomes, but also in loyalty).

More conversations from the Summit coming, but let me know your thoughts on this one.  Do you agree?  Disagree?

Christina Phillips is a Microsoft Certified Trainer and Dynamics GP Certified Professional. She is a director with BKD Technologies, providing training, support, and project management services to new and existing Microsoft Dynamics customers. This blog represents her views only, not those of her employer.

Thursday, October 6, 2016

How do you choose your passwords and keep them secure?

By Steve Endow

I've had a mini discussion on Twitter recently about passwords.  There was some joking about the stereotype of the password written on a Post-It Note stuck on a monitor, and about how common passwords tend to be re-used.

I noted that I preferred P-Touch labels for my monitor passwords, as they worked better than Post-It Notes.

I occasionally have calls with Dynamics partners who store all client information in a central system, such as MS CRM Online.  In addition to the standard customer information, they store VPN connection information, Windows logins, and SQL Server sa passwords in plain text in the CRM customer records.  Logistically this makes sense, as it allows all of the consultants to quickly access login info and assist customers.  But it poses a potential security risk, as a single compromised CRM login could expose all customer connection information, and full access to customer Dynamics GP SQL Servers and databases.

Monday, October 3, 2016

Why Care? A Case for Really Understanding Your Account Framework

It seems like most of us have had a time in our careers as consultants where we did it all.  We installed the software, configured and trained, migrated data, and even did an upgrade or two.  Over the years, my job has evolved where now I spend a lot more time working with a team.  I don't remember the last upgrade I did (maybe three or four years ago?), other than on my own machine.  And I don't recall the last new install I did. 

So it occurs to me that as we bring people in to consulting, the "do everything" approach is not really used much anymore from what I can tell.  Due to the increasing technical complexities of installing and supporting systems, and larger partner organizations, we tend to see people fall in to one or two of three primary categories- functional, technical, developer.  And for this reason, there is sometime a disconnect between things that each group should know and understand. 

Many years ago, I made the case for waiting to install the software until after we had completed our discovery.  It was common practice to schedule installation as soon as we had registration keys, which sometimes meant even before we had formally kicked off the project.  Why is this an issue?  Account Framework.  This thing that isn't an issue very often, but when it is-- it is a significant issue that can lead a customer to say "Why didn't you anticipate this?".  Which, by the way, is something project managers don't want to be asked.

I want to make a case that everyone on your team, and clients/users as well, needs to understand the significance of the Account Framework.  In many ways, it is the most critical decision (or lack of making a decision) encountered during the installation process.  Handled poorly, it can cost you money, time, and project goodwill.  At a high level, the Account Framework is comprised of three components- maximum number of segments, maximum number of characters for the entire account string, and individual maximum characters for each individual segment.

The Account Framework is defined during the creation of the Dynamics GP system database.  It is defined at this stage because it impacts the fields and sorting available for the chart of accounts in ALL COMPANY DATABASES.  Did you catch that?  ALL COMPANY DATABASES.  Now, let's contrast with the Account Format.  The Account Format is defined during the actual configuration (logging in and setting up) in each company.  The key here is that the Account Format can be different in each company, but all companies' Account Formats must be able to individually fit within the overall Account Framework?


Well, let's take the standard example of 66 characters and 10 segments.  This is the largest possible Account Framework in Dynamics GP.  This allows us to grow to use 10 segments of 6-7 characters each (as it allocates the 66 over 10 segments).  So the following Account Formats would be possible with a "66 and 10" Account Framework:


Wouldn't be possible in this example?




It is important to keep in mind that the Account Framework is the largest the installation can GROW.  So maybe your initial account format for a company is XX-XXXX-X so you decide to specify the same as the Account Framework (a practice we come across frequently).  This works great initially, but if the client ever wants to expand their Account Format (a fairly common request when someone is on the software for a number of years, or even sometimes in the course of an implementation when the need to add an additional segment comes to light), they will have to chance the Account Framework (more on that later) which is not the flip of a switch.  Anecdotally, I have not seen evidence that a smaller account framework yields performance or other benefits.  But if you are reading this and you have seen such benefits, please share!

When a client wants to change the Account Format, it is a fairly straightforward process.  Then can update the Account Format Setup (Setup-Company-Account Format) and then as needed use the PSTL Account Modifier/Combiner tool to update segments as needed.  But if they encounter the boundaries of the Account Framework when doing this, it becomes a much larger issue.  Technically, you can script the changes to the Account Framework to alter the necessary tables and components.  But the recommended approach would be to use the Reformatter tool from Corporate Renaissance (which effectively creates a new Dynamics database and ports over the needed data)...

But of course that involved buying a product, testing, and then deploying the fix.  Microsoft used to also offer a service to do this through support (for an additional fee), but that involves additional down time as the data is sent to MS and returned.  In all cases, additional cost and time is involved.  And it takes a change that is intended to make the system flexible (e.g., being able to change the account format and use the PSTL account modified/combiner) and user-friendly, and adds cost and potential consulting hours.

In my not so humble opinion, before an install happens, everyone on the project team should understand what framework is being requested and why.  If you want to check your account framework, one of the easiest places to do so is Setup-Company-Account Format where you can see the maximum number of segments, characters, and the max length for each segment per the framework.

For reference, when installing Dynamics GP for the first time, GP utilities will allow you to select either a Basic or Advanced installation.  The Basic installation will install with a default Account Framework of maximum 5 segments, 45 characters, and 9 characters per segment.  This cannot be changed.  An advanced installation will allow you to specify the framework.

Christina Phillips is a Microsoft Certified Trainer and Dynamics GP Certified Professional. She is a director with BKD Technologies, providing training, support, and project management services to new and existing Microsoft Dynamics customers. This blog represents her views only, not those of her employer.

Sunday, October 2, 2016

Supporting .NET 2.0 references in a 4.0+ class library at runtime

By Steve Endow

If you don't know what a .NET mixed mode assembly is, you can skip this post and save a few minutes of your life.  If you do know what a .NET mixed mode assembly is and understand why it can be a pain, you may want to read on.

This post is primarily to document this solution for myself so that I can find it again in the future without having to dig through 10 old projects (like I had to do yesterday).

This solution is something my colleague Andrew Dean and I have shared during our Dynamics GP .NET Development presentations at the reIMAGINE and GPUG Summit, as it is something we have had to use in our development projects with Dynamics GP.

In the case of Dynamics GP, the primary reason I've had to use this is because I regularly use GPConnNet.dll.  That dll is still only .NET 2.0 'compatible' (or whatever the term is).  So if you develop a new solution using .NET 4.5.1, you will get the famous mixed mode assembly error.

Mixed mode assembly is built against version 'v2.0.50727' of the runtime and cannot be loaded in the 4.0 runtime without additional configuration information.

The standard solution to resolve this error is to modify your application config file.  Meaning your app.config / exe.config file.

That works fine if you are developing an EXE, but it won't work if you are developing a class library DLL that will be used by other applications that you can't control.  Like Dynamics GP.

This post by Reed Copsey Jr. provides an alternative, and brilliant, approach to setting the .NET legacy policy at runtime.  This is a life saver for DLL libraries that are referencing resources with multiple .NET versions.

public static class RuntimePolicyHelper
    public static bool LegacyV2RuntimeEnabledSuccessfully { get; private set; }

    static RuntimePolicyHelper()
        ICLRRuntimeInfo clrRuntimeInfo =
            LegacyV2RuntimeEnabledSuccessfully = true;
        catch (COMException)
            // This occurs with an HRESULT meaning 
            // "A different runtime was already bound to the legacy CLR version 2 activation policy."
            LegacyV2RuntimeEnabledSuccessfully = false;

    private interface ICLRRuntimeInfo
        void xGetVersionString();
        void xGetRuntimeDirectory();
        void xIsLoaded();
        void xIsLoadable();
        void xLoadErrorString();
        void xLoadLibrary();
        void xGetProcAddress();
        void xGetInterface();
        void xSetDefaultStartupFlags();
        void xGetDefaultStartupFlags();

        [MethodImpl(MethodImplOptions.InternalCall, MethodCodeType = MethodCodeType.Runtime)]
        void BindAsLegacyV2Runtime();

Go forth and develop mixed mode assembly class libraries.

You can also find him on Google+ and Twitter

Saturday, October 1, 2016

Cool Visual Studio Immediate Window trick

By Steve Endow

I'm a big fan of the Immediate Window in Visual Studio.  When you are debugging code and trying to figure out a problem, the immediate window is a simple and quick way to check variable values, test operations at a breakpoint, and take a look at what is going on.

But there is one small issue.  If you are outputting lots of text, such as an eConnect XML document or eConnect exception message, the Immediate Window will output the entire XML document as a single line of text.  A really, really long line.  Worse, newlines will be displayed in the text as \n\r.  

This makes outputting XML to the Immediate Window a bit of a hassle--I have been copying the text, pasting it into UltraEdit, replacing the newline characters with actual CRLF line breaks, and then formatting the XML.  It's a hassle.

So, once again, my laziness to repeat that process finally overcame my laziness to actually do some research and see if there was a better solution.  Of course, there is a better way.  I just didn't know it for the last TEN YEARS.  I'm apparently a slow learner.

Big surprise, the answer was on Stack Overflow.

The trick is to type ",nq" after your Immediate Window command to request "no quotes".  Apparently this is a "format specifier".

Holy cow, what a huge improvement.  It might seem trivial, but it makes a huge difference.  Here's an example of before and after.

Notice the standard "? response" command outputs that variable text (an eConnect exception) as a single line that's a pain to read.  Once I typed "? response,nq", like magic, it output the text in a very nicely formatted manner.  Brilliant.  Soooo much more readable and it saves me a half dozen useless steps in a text editor.

There are at least 1 billion shortcuts and hidden gems in Visual Studio that I still don't know, but I'm finding them one at a time...  Pacing myself.

You can also find him on Google+ and Twitter