Archive for category Uncategorized
I recently upgraded a Azure Cloud Service from the 2.3 SDK to the 2.5 SDK. I keep the IIS log files for analysis but after the upgrade I noticed that I stopped getting the files uploaded to the storage account.
Digging around I found this was an issue with the Azure 2.5 SDK and that the workaround is to modify the OnStart method in your WebRole.cs file. This requires a restart of your app before you’ll start seeing the IIS log files copied over to your storage account.
One thing that I didn’t expect is that the Azure Diagnostics only copied newer files. I guess this makes sense, so I went digging around for a way to copy the older files to the storage account.
Here’s what I came up with.
Connect via Remote Desktop to the Web Role in Server Explorer.
Open a Poweshell prompt and download AzCopy.
PS> invoke-webrequest http://aka.ms/downloadazcopy -OutFile AzCopy.msi
Run the AzCopy.msi installer and open the Microsoft Azure Storage Command Line. This opens a command window with the path to the azcopy executable already in the path.
Now before I could go further I wanted to place the log files in the same directory structure that Azure Diagnostics already supports. Querying the storage account I see they are located in the following location:
But where are the files located on the Azure Cloud virtual machine? Digging around I see that the log files are stored at the location
To make this easier to see on the command line, I created a configuration file with all the arguments I needed to pass to AzCopy. Here’s my configuration file (excluding private information)
/Source:C:\Resources\Directory\<deployment_id>.<Site_Name>.DiagnosticStore\LogFiles\Web /Dest:https://<account>.blob.core.windows.net/wad-iislogfiles/WAD/<deployment_id>/<Site_Name>_IN_<Instance>/ /DestKey:<storage account key> /S /BlobType:block /V
Knowing this I could now use AzCopy to copy the old IIS log files up to the storage account using the following AzCopy command.
After AzCopy was done copying the files, I repeated the same operations on the other nodes in the Azure Cloud Service and verified the files were there by using the BlobTransferUtility, that I use to copy the files locally.
It’s been a long time I last updated.
When I last posted I was lining up where the front cross-member and the brackets for the 1985 Corvette suspension. On the chassis table I need to build what I’m calling “stations”. These are locations along the length of the frame where specific attachments would be bolted to the frame. These attachments are things like the front & rear bumpers, running board brackets. These stations help locate the frame on the chassis table and keep everything lined up, in addition to building additional stations to help locate the front cross member and the brackets to mount the independent rear suspension of the ’85 Corvette.
I was working on building the stations for the front and rear running board brackets. The flat bar has this hard layer to keep the steel from rusting. It’s a pain to remove and even with 40 grid flapper disk, it takes a lot of effort to remove. I also noticed that the suspension brackets were developing some surface rust. I wanted a sand blast cabinet.
I went online to TP Tools and really liked their cabinets, but after starting with the basic model and upgrading to the next one “because it’s only a few hundred more”, when I was done I was looking at a $3500 cabinet. Years ago my dad built one out of 18ga but I gave it to my dad’s lifelong friend when he passed away. Because I’m my fathers’ son, I decided that I’d build my own.
Off to Discount Steel to pick up some 16ga for the main body, 18ga for the hopper, 11ga for the doors and 1.5″ x 1.5″ x 11ga angle iron to build the work support framework.
I started off with trying to build the hopper with the 18ga sheet. It was easy enough to cut out with the nibbler but then I tried to bend it. I thought I could just clamp it and bend it over. This didn’t work, so then out came the body hammer and I tried to bend a 1″ flap over 24″. This turned into an absolute mess. I just wasn’t happy with the results.
After looking online for a sheet metal brake that could bend 16ga sheet over 5′ long I came up with the brain-dead idea that I could build my own sheet metal brake.
After screwing around with this idea for a month I was severely unmotivated and was stuck like this for several months. I didn’t get back to the shop until August.
Over the next few posts I’ll describe the progress I’ve made with the sand blast cabinet.
It’s been a while since I posted but I wanted to catch everyone up to where I’m at in the build process.
When I started off writing this post I was going to go into detail about what’s happened over the last 6 months or so, but this just became too much work.
Suffice to say I’m starting over.
What happened was that I cut built my own frame by cutting pieces out of 1/8″ steel plate using my plasma cutter. Thing were going along well until I put too much heat into the welds and ended up warping the frame rails. I thought I could straighten them out with some heat from the oxy torch but not to my satisfaction, or more likely, I don’t have the necessary skills to do this.
So I’ve decided to start over, but this time I’m going to put a little more thought into it.
Over the last few months I’ve been corresponding with a fellow car nut from Ireland who is also interested in building a ’56 Ford F-100. Problem is, he doesn’t have access to an original frame. Ian turned me on to a builder Paul Horton ( www.welderseries.com ) who has an interesting way of making parts for your hotrod.
These little tabs on each piece help locate the parts relative to each other. Kind of like the paper models I used to put together where you put Tab A into Slot B.
Taking inspiration from this approach I set about making some CAD drawing of the ’56 F100 frame I do have. I started off tracing the outline of the frame onto a sheet of 1/4″ 4′ x 8′ plywood. Since the frame is almost 16′ long I had to trace it as a front and rear section. From this outline I could start taking dimensions to holes that I needed to keep. I’d center punch the hole into the plywood and then take measurements from this indentation. What was really a challenge was all the curves on the frame. I really struggled with this until I remember some high school geometry around tangents and circle radius being perpendicular to the circumference. What I did marked every frame transition from a flat part into a curved part. At this transition, I would draw a perpendicular line or ray. When the curve stopped, I would draw another perpendicular line or ray. Where these two lines met, that would be the centerpoint of the arc. I could then measure this distance and the angle to know how to draw this arc in the TurboCAD.
Here you can see what the plywood ended up looking like when I was through (it’s hard to read all that’s there):
From this tracing I started transfering the dimensions into TurboCAD. After spending several weeks doing it over and over again I came out with something that I felt was close enough to what I needed. And after checking with a local steel company (www.discountsteel.com) around how much it would cost to have this cut out, I sent the plans off.
Today I picked up the parts and boy does this look fantastic. Here are some pictures showing how the front and rear sections will go tother as well as a few pictures for other brackets that are used on the frame.
For some reason the rear section messed up. I ended up with two inside parts instead of an inside/outside part. I checked the drawing I sent and both pieces are there. I’m not sure how this happened. I’ll have to check with the DiscountSteel to see if they had to make any modifications to my drawing.
Here you can see how the front and rear sections will be joined together. I’d be hard pressed to be able to produce this accurate of a cut using the plasma cutter by hand.
The main body mount to the frame. I’ll go into details on why I made this as 5 pieces instead of one in a future post.