Calling iSeries DB2 web services from .Net

Share

Problem

Instead of giving you direct access to make DB2 calls to your corporate iSeries, your IT department is exposing iSeries database capabilities via web services provided by IBM’s WebSphere.  Sounds great, but as you’ve experienced, it’s a major pain because they’re not implemented the way you expect.

Solution

  1. Get the URL from your iSeries team for their web service.
  2. In your .Net project, right click “References” or “Service References” and choose “Add Service Reference”.
  3. In the “Add Service Reference” dialog, enter the URL they provided to you into the “Address:” field.  Be sure you add “?wsdl” to the end of it if it’s not already there then click “Go”.
  4. You might be prompted for credentials.  Be sure to get those credentials from your iSeries team.  You will likely be prompted to enter them 3 or more times.  Yes, it’s a nuisance, but just do it.

Now you’ve got the web services added, but you’ve got some config file editing to do.

In your app.config or web.config file, find your custom binding for this service.  If you had to enter credentials, you’ll need to change the httpTransport to what you see below, but the realm will be different.  Get that from your iSeries team.

<customBinding>
  <binding name="Some_Crappy_NameServicesPortBinding">
    <textMessageEncoding messageVersion="Soap12" />
    <httpTransport authenticationScheme="Basic" realm="Secure_SOMETHING" />
  </binding>
</customBinding>

If your iSeries requires credentials, you’ll need to set them on your web proxy like this before you call a method:

ClientCredentials.UserName.UserName = username;
ClientCredentials.UserName.Password = password;

Now, to call a method on the web service is quite different.  You’ll have 4 classes provided in your proxy:

  1. Input
  2. Request
  3. Response
  4. Result

Each one has a name prefixed with “fweb” and some number.  For example, “fweb12Input”.  Each web service your iSeries team adds will have a new number.  Yes, this is entirely backasswords and highly inconvenient, but that’s the way IBM has done it.

You’ll want to instantiate a request object.  It has a field called “arg0” in it.  You’ll want to assign that to a newly instantiated Input object.  The Input object has fields in it representing what would normally be parameters to a web method.  Here’s an example:

var request = new fwebr024Request
{
    arg0 = new fwebr024Input
    {
        IN_FIRSTNAME = "John",
        IN_LASTNAME = "Smith",
        IN_SESSIONID = "Whatever",
        IN_USERID = "DOEJANE"
    }
};

Then you’ll call the web service like this:

var result = this.MyServiceProxy.fwebr024(request);

The result object will have the output of the web service.  It has a strangely named field called “@return”, which is an object with 3 fields representing any error that might have occured:

  1. OUT_ERRCODE
  2. OUT_ERRSTATE
  3. OUT_ERRTEXT

That’s it.  It’s pretty harry, but that’s how you do it.

See these images?

image

You’ll find actual working versions of them at the top and bottom of this article. Please click the appropriate buttons in it to let your friends know about this article.

Check back later for updates too!

Share

Add a web.config transform and a publish profile in Visual Studio 2013

Share

If you need to deploy your app to multiple environments, like in most corporate IT shops (dev, QA, Staging, Training, Production, etc…), then you’ll need to have multiple versions of your config files, or better yet, one config file and “transform” files (one for each deployment environment) that describe ONLY the differences between the main config file and that particular environment.

For example, the connection string in the QA environment is likely different than the connection string in your Dev environment, which is different than your QA environment and again different in your live Production environment.

I’m not going to explain how to WRITE a config transform (how to tell it what needs to change).  At least, not in THIS article.  But I will tell you how to tell Visual Studio that you have multiple environments and how to make Visual Studio create the basic config transforms for you.

In this example, I’m creating a WCF Service application (works the same with pretty much any web type of application).

  1. Are you deploying a NON Asp.Net app (like a click-once app or a WinForms or WPF app)?  If so, install the Nuget package “Slow Cheetah”.  Why?  Because Visual Studio has built in support for all this for web.config files, but NOT for app.config files.  Slow Cheetah lets you make transforms for ANY file in your project.
  2. Right-Click your project and choose “Publish…”transform_01
  3. In the “Publish Web” dialog, choose “Custom”transform_02
  4. Give it a name.  NOTE!  If you have a different admin managing your deployment and/or build servers, you may want to check with them on what name to use, because it will make a difference between whether your stuff works or doesn’t!  For this example, I’ll call the transform “QA”transform_03
  5. Choose your deployment method (web publish, file copy, etc…).  For this example, I’m choosing “File System” since it requires fewer settings to fill out and I’m going to leave “Target location:” blank.  My deployment admin will fill that in later, so I don’t even need to know this.  Click “Next”.
  6. Choose whether this deployment should be a “Release” or a “Debug” deployment.  This will cause it to build it as debug or release. (You will also have a debug and a release transform of your web.config file and this new QA transform will inherit from either of those).transform_04
    1. Expand “File Publish Options” and check the items you need, then click “Next”
  7. Final screen in the wizard.  Click “Close”.  You can’t click “Publish” if you left the target path blank above.

You’ve now successfully created a publish profile.

transform_05 QA done

Now you’ll need to create a Web.config transform for this profile.

  1. Rich-Click your QA.pubxml file and choose “Add Config Transform”.  Do NOT choose “Add Transform” if you have Slow Cheetah installed.transform_06 QA Add Transform

You now have a new Web.QA.config file.

transform_07 Complete

You can now code your base Web.config file the way you need it to run locally during development.  In your Web.QA.config file, you can add transforms to modify settings in your web.config file so that when you build for that environment, Visual Studio will produce a web.config file that’s right for that environment.

You can repeat these steps to add as many publish profiles as you need.

See these images?

image

You’ll find actual working versions of them at the top and bottom of this article. Please click the appropriate buttons in it to let your friends know about this article.

Check back later for updates too!

 

 

 

Share

My Love/Hate Relationship with Amazon.com and a serious flaw in their service

Share

I love Amazon.com.  I love their selection.  I love their prices.  I love their return policies.  I love my Prime membership.  I love their delivery options and prices and speed.

But, Amazon.com has a nasty, horrible flaw.   And it’s horrible on multiple levels.  Let me explain the flaw and then how it’s exacerbated by them:

The Technical Flaw

For the past week or so, I’ve been having trouble logging into my Amazon.com account.   It started with an e-mail message I got that looks exactly like a phishing email message… you know the kind… That warns you that someone might have been messing with your account so they had to lock your account and that you need to click this link in the email to log in to “verify” you account?

Yeah, that’s what I got.  But all the links in the e-mail went to the actual Amazon.com site, which was unusual, because everything else about the e-mail threw up ALL the red flags of a phishing scam, so I simply ignored it.  But, no less than a day later, I started having trouble logging in.  I KNOW my login credentials I was using were correct because they’re stored in LastPass and LastPass is the one that enters them into the form.  Also, the Amazon app on my phone suddenly couldn’t log in either.

So, I went through the “I forgot my password” routine, had some e-mail sent to me with a link to change my password.  I changed it, but I STILL couldn’t log in.  I changed it again and I STILL couldn’t log in.  I changed it again and I STILL couldn’t log in.  I tried the latest, newest PW on my phone and I STILL couldn’t log in.  I tried it from Chrome, from FireFox, from the app on the phone, from Chrome Beta on the phone… all with the same failure.

I called tech support and after an hour and 14 minutes, talking to 2 agents, I was able to confirm (well, I was able to prove to them) that the problem was indeed on their end.

The problem is the CAPTCHA image they’re displaying in my browser for me to read and type in the letters is a DIFFERENT captcha than what their server thinks it sent me.  For example, if the image they send has the text VRB7TC in it, their server THINKS it sent me an image with KFB98RX in it, so when I type the right text, it still thinks it’s wrong.  But, somewhere between 2 to 7 tries, it eventually gets it right.  But if I log out and back in again, I go through the same nightmare.

How do I know it’s a CAPTCHA problem on their end?

Two reasons:

  1. While on tech support, the agent read to me the CAPTCHA text that their server said it sent me.  It was NOTHING like the captcha text on my screen.
  2. The agent finally admitted this is a problem they know about.

They actually willfully make this problem worse

  1. They wasted my time and they wasted two of their agents time by willfully lying to me, giving me the following lies:
    1. It’s my fault for logging in and out so often, causing my account to get locked out (BTW, it was never locked out)
    2. I need to upgrade my browser (I have the latest version of Chrome and FireFox on Linux and Chrome Beta on Android AND their own app on Android (also latest version) failed too).
    3. Told me I need to clear my cache.
    4. Told me it’s because of cookies that I need to delete.
    5. Telling me (I’m a web developer, just so you know) that even though I’m typing in the right password, my browser is sending the wrong one.  (This gets 7 out of 4 Pinocchios… Yes, you read that right).  Yes, I educated them on how this works.
  2. Not admitting early on that this is a known problem.

It gets worse!

After I berated the second level agent that all of his excuses were bullshit and explained in precise detail WHY they were, he finally admitted that this is a KNOWN PROBLEM!  This is AFTER he said he opened a ticket for me.

A KNOWN PROBLEM???  WTF didn’t he tell me that at 1 minute into the call?  Why drag me along for over an hour?

I asked when I should expect it to work for me.  He responded with gibberish and no answer.

I asked again, “When will this be fixed for me”.  He reordered the words of his gibberish with no answer.

I asked a third time, “But when will this be fixed for me?”.  Again, gibberish.

I asked a fourth time, “When will this get fixed for me?”.  This time he said, “Soon.  They’re actively working on it”.

My next obvious question was, “WHEN?”.  Him:  “Soon”.

Me:  “How long have you known about this problem?”

Him:  “Well, it was caused by security updates in your browser”.

Me:  “How long have you known about this problem?”

Him:  “It’s when the browsers made a security change that caused this to start happenning.”

Me:  “How long ago did this start?  I’m concerned that you’re not giving this the attention it needs.”

Him:  “We found out about it in October 2014”

Me:  “WHAT?!?!?!  You’ve known about this for SIX MONTHS?!?!?!?  And it’s STILL not fixed?  This is UNNACCEPTABLE for a company as big as Amazon.  This is CRITICAL and you should have had programmers working on it as a priority one and should have had it fixed on DAY ONE!  This is the CORE of your business!  If we can’t log in, we can’t buy anything!”

The call ended politely and he acknowledged that it was not a good situation, but I was left with ZERO resolution.  Nothing changed.  It’s still broken, and they’re not really going to do anything about it.

THAT’S what’s so insane and inexcusable about it.

Before you shout, “BUT IT WORKS JUST FINE FOR ME!”, let’s make perfectly clear, this is not a global problem.  It happens on some accounts.  Unfortunately, mine is one of them, and they have no idea why and don’t appear to be doing anything about it.

The Silver Lining and tiny shred of hope

The agent told me they are beta testing 2 factor authentication (though the agent didn’t know that’s what it’s called and kept calling it “a way for you to enter a code when logging in”.  After interrogating him, I was able to determine it is actually 2-factor authentication.

He showed me where, on the website, to activate it, but sadly, it was not available to me.  I asked him to add me to the beta.  He said he can’t and that it is just rolling out to random people over a period of time.

If you’re interested, this is where you can find it:

  1. Go to your Account page.
  2. Scroll down to Settings.
  3. Underneath “Forgot your password” and above “1-Click Settings”, there will be a new item.  I don’t know the text, but it will be something like “2 factor authentication”.

 

Amazon 2 factor

If you don’t see it, that means you haven’t been given the roll out of the newer version of their website, and no, you cannot call up and ask to be put on it.  Trust me, I tried.  They can’t do it.

I find it amazing that Amazon could let such a serious flaw go on for half a year (so far).  This is the kind of bug you give top priority to and put programmers on it until it is completely resolved.

See these images?

image

You’ll find actual working versions of them at the top and bottom of this article. Please click the appropriate buttons in it to let your friends know about this article.

Check back later for updates too!

Share