My original thinking, is that if you don’t have any conditions, it would just route every request that matches the incoming URL, this doesn’t seem to be the case.

Make sure you have at least one rule {HTTP_HOST} or something that will be the catch for the requests coming in.

2011-11-07T15:16:28.8528558+00:00 Error ITSM & Farm Hq-aesop0 Failed to run operation ‘RunRemote’.  Failed to run method ‘Microsoft.Web.Farm.GetInstalledProductsRemoteMethod’ on server ‘Hq-aesop0′.  Exception in response stream.  InternetOpenUrl returned 0x80072EFD: Unknown error 12029.  Exception in response stream.  InternetOpenUrl returned 0x80072EFD: Unknown error 12029. Microsoft.Web.Farm.WebFarmException: Failed to run method ‘Microsoft.Web.Farm.GetInstalledProductsRemoteMethod’ on server ‘Hq-aesop0′.  Exception in response stream.  InternetOpenUrl returned 0x80072EFD: Unknown error 12029. —> Microsoft.Web.Farm.WebFarmException: Exception in response stream —> System.Net.WebException: InternetOpenUrl returned 0x80072EFD: Unknown error 12029.

In WFF 2.2, you can set a proxy server that all your servers will use in your applicationHost.config. This setting will be passed to all your servers for their calls out to the Web Products Feed xml file used by Web Platform Installer. You can set this using the following appcmd:

appcmd.exe set config -section:webFarms /[name='Primary'].defaultProxy.enabled:"True" /[name='Primary'].defaultProxy.proxyaddress:"proxyservername:80" /[name='Primary'].defaultProxy.userName:"testuser" /[name='Primary'].defaultProxy.password:"Password1" /commit:apphost

You will need to replace the name of your webfarm where it says ‘Primary’ and put your proxy details where it says "proxyservername:80". If you proxy requires authentication, replace "testuser" and "Password1" with the username and password respectively.

vertarget

Show Threads

!threads

Switch Threads

~IDs

Dump the current heap

!dumpheap [-stat for a summary]

Dump Object

!do 0xXXXXXX

Get the URLs in the request objects in memory:

.foreach (req {!DumpHeap -type System.Web.HttpRequest -short}) { .foreach /pS 1 (a {dd ${req}+48  l 1}) {.echo ${a}; !do poi(${a}+c)} }

Print exceptions in memory:

.foreach(exception {!DumpHeap -type Exception -short}) {!PrintException exception; .echo print exception done !!! *****************}

When I start work with a new customer, it is up to me to extract what the real problem is. Be it a broken system, a bit of software they are writing, or some consultation they need before starting work on a project. The key to this is communicating clearly with the person at the other end of the phone.

There is some key information that is always useful in a business critical situation. The first of these, and most obvious, is what the actual problem they are facing is. This can range from a very simplistic issue, up to a massive deployment, spanning multiple server farms and technologies. Secondly, it is good to understand what impact this is having on their business. This allows you to gauge how important this issue needs to be dealt with. I have helped customers in cases where they are losing hundreds of thousands of pounds every hour there system is down. Getting this information upfront sets the tone of your work, and makes you realise just how important you are to a company, which, as an intern is quite a shock at times!

If you are ever in a situation where by you are required to handle work over the phone, and this could include being interviewed (thinking about the Intern & MACH application process!) then I can’t stress how helpful it is to have a plan in front of you before you even pick up a phone. Make a note of all the topics you want to cover, any questions you need to ask, the name of the person you are calling (I am forever forgetting the name of the person I am talking to…in real life as well!), and any other points you need to bring up. There is nothing worse than hanging up the phone and remembering a question you needed to ask.

When I was at the telephone interview stage of my application to the Microsoft Intern Scheme, I sat down the day before my scheduled interview time, fired up the trusty OneNote (best noting taking programming in my opinion – it comes part of Microsoft Office….look in All Programs > Microsoft Office > OneNote) and noted down all the points that I thought might come up and any key topics that I wanted to make sure I mentioned. Thankfully it seemed to work!

Through careful preparation, even the most daunting call can be a stress free experience.

One method of troubleshooting this is to enable something called Event Tracing for Windows, specifically tracing the events from the HTTP.sys system. HTTP.sys is the kernel level side of a web request. All requests that go to IIS pass through this, so it will trap any events.

To start tracing events in Http.sys use the logman.exe tool. Logman.exe ships with every Windows installation, and can be used right out of the box. To enable ETW, type:

C:> logman start httptrace -p Microsoft-Windows-HttpService 0xFFFF -o tracefile.etl -ets

Here tracefile is the name of our trace. This will be used to identify this particular event trace in further commands. Logman needs to know which components of Windows to trace. Since we are only interested in tracing Http.sys, we pass the provider name, "Microsoft-Windows-HttpService". The 0xFFFF signifies we want full tracing, or all trace events. Our event trace will be written to a raw binary file, tracefile.etl. Please note that you must be an elevated Admin to use the logman command.

To stop tracing we use the logman tool again:

C:> logman stop httptrace –ets

You can now open up this ETL file up in Event Viewer, which give you a very nice verbose view of the request passing through.

Now you will be able to see if a request is passing through your system or not!

Feel free to email me alex[at]olivierdigital[dot]com if you need any help!

Alex

I have been struggling to get to grips with this stuff, heaps, stacks, clrs, gc’s, gen 0, etc, and have finally come across a very useful duo of blogs posts that cover how to get setup and start finding the bugs:

Getting started with windbg – Part I and Part 2

These are written by JohanSt over at MSDN. I can also recommend reading through my colleague Doug Stewart’s blog who is the Principal Escalation Engineer in my team, and trust me, this guy is a genius/Jedi/Matrix-dweller.

<script>document.body.onselectstart=function(){return false}</script>

Simple as, put that just before you </body> tag, or call it onload, and the default event is overridden and no highlighting occurs.

Case closed.

Microsoft UK is now open for applications for the Graduate and Intern Schemes for 2012. This really was the best decisions I have ever made. I have met so many cool people, done some amazing things, and I’ve only been here 2 months!

Apply now!!

Its really simple to apply, all you have todo is fill in a form, and make a quick video explaining why you want to be an MSP. Full details can be seen here.

If you have any questions about it feel free to get in touch with me on Twitter, @alexolivier, or contact Ben Nunney @bennuk who runs the program here in the UK.

Just as a bit of fun, below if a video I did when I went to my first MSP event last winter: