mystyleit.com

Finding Messages inside Exchange using EntryID

mike.clarke — Tue, 13 Jul 2010 00:05:00 GMT

Finding Messages inside Exchange using EntryID

Recently I have had a need to locate messages inside Exchange using only the EntryID. Ideally I wanted to do this through Outlook but all evidence points towards this not being possible.

http://www.officekb.com/Uwe/Forum.aspx/outlook-prog-forms/571/EntryId
Outlook doesn't export that property. Sue Mosher, Outlook MVP
http://help.lockergnome.com/office/Find-EntryID--ftopict937618.html
Never mind on the EntryId I found where it is not valid for the Find command.

The only thing I could find was to use Outlook shortcuts/hyperlinks:
http://www.slipstick.com/outlook/links.htm
But these proved very unreliable. In fact I only managed it make it work once.

Other Possibilities

Based on everything I read, the only way to find the message is through API/SDK. Specifically using GetItemFromID:
http://msdn.microsoft.com/en-us/library/bb219902%28office.12%29.aspx

Solution

My solution was to write a really small C# command line that app that will open the message based on EntryID and StoreID.

using System; using System.Collections.Generic; using System.Text; using Microsoft.Office.Interop.Outlook; namespace EntryID { class Program { static void Main(string[] args) { Console.WriteLine(""); Console.WriteLine("REQUIRES OUTLOOK"); Console.WriteLine(""); Console.WriteLine("THIS APPLICATION COMES WITH NO WARRENTY OR GAURENTEES"); Console.WriteLine("USE AT YOUR OWN RISK"); if (args.Length != 2){ Console.WriteLine("ERROR: Missing Parameters"); Console.WriteLine("EntryID.exe storeID entryID"); Environment.Exit(-1); } string storeID = args[0]; string entryID = args[1]; Console.WriteLine( "storeID: " + storeID ); Console.WriteLine(""); Console.WriteLine( "entryID: " + entryID ); Microsoft.Office.Interop.Outlook._Application olApp = new Microsoft.Office.Interop.Outlook.ApplicationClass(); Microsoft.Office.Interop.Outlook._NameSpace olNS = olApp.GetNamespace("MAPI"); Microsoft.Office.Interop.Outlook.MAPIFolder oFolder = olNS.GetDefaultFolder(Microsoft.Office.Interop.Outlook.OlDefaultFolders.olFolderInbox); Microsoft.Office.Interop.Outlook._MailItem oMailItem = (Microsoft.Office.Interop.Outlook._MailItem)olNS.GetItemFromID(entryID, storeID); oMailItem.Display(false); } } }

The hardest part was getting my Visual Studio app to hook into Outlook 2003/Outlook 2007. The following MSDN articles were very helpful:

Requirements

In order to use this application, you must have the Office Primary Interop Assemblies (PIA) installed for your version of Outlook.

I do not believe that Office 2003 ships with PIA, see Hello Word Outlook Add-In using C# but I believe Office 2007 does.

You can download the PIA for Office 2003 and Office 2007 using the following links:

Here is a link about doing development with PIAs
http://blogs.msdn.com/b/vsto/archive/2008/05/20/common-pitfalls-during-pia-deployment-and-installation.aspx

Results

I ended up making two apps. One for Outlook 2003 and one for Outlook 2007. From the command line I can now open a message using the storeID and entryID. Here is the prototype for app: EntryID2003.exe storeID entryID EntryID2007.exe storeID entryID

When you run the app with the right command line parameters, the app will start Outlook and open the email.

The compiled versions of the app (both Outlook 2003 and Outlook 2007) can be downloaded here:
EntryID.zip

Update: Finding Messages inside Exchange using EntryID

After continue working on this I had problems getting EntryID2003.exe working on Outlook 2003. I believe this is a problem linking into Outlook. Since then I have come up with a new solution in the form of a VB script.

Here is the script

'************************************************************************* 'THIS APPLICATION COMES WITH NO WARRENTY OR GAURENTEES 'USE AT YOUR OWN RISK '************************************************************************* 'Instructions 'Fill in the parameters below 'Close Outlook and the run the script. This script should open the message with EntryID. '************************************************************************* 'PARAMETERS storeID = "0000000038A1BB1005E5101AA1BB08002B2A56C20000454D534D44422E444C4C00000000000000001B55FA20AA6611CD9BC800AA002FC45A0C0000004443002F6F3D4669727374204F7267616E697A6174696F6E2F6F753D45786368616E67652041646D696E6973747261746976652047726F7570202846594449424F484632335350444C54292F636E3D526563697069656E74732F636E3D6176616E64616C617900" entryID = "000000007AA33668D54A1C4D9D8372FEE1EC794F07007D2133630C6123449784F16966F749C0000009264F2F00007D2133630C6123449784F16966F749C00459A9A964650000" 'Set this to the value of your Outlook profile profile = "Outlook" Set objOutlook = CreateObject("Outlook.Application") Set objNamespace = objOutlook.GetNamespace("MAPI") objNamespace.Logon profile,, False, True Set message = objNamespace.GetItemFromID( entryID, storeID ) message.Display

The script can be downloaded here:
EntryIDScript.zip

Xbox 360 Musings

mike.clarke — Mon, 28 Jun 2010 23:22:00 GMT

Xbox 360 Musings

I recently bought a used Xbox 360 and have been messing around with it for the last couple weeks. Since then I've learned some things and thought I would put them all in one place.

Original Xbox Support and System Update 9199

The Xbox 360 can play some Xbox 360 games. The official list has been taken down by MS. Here is the unofficial list:
http://en.wikipedia.org/wiki/List_of_Xbox_games_compatible_with_Xbox_360

The real trick is getting it to actually work. It seems like nearly every link on the Internet points to this update:
http://www.xbox.com/en-US/support/systemuse/xbox360/console/systemupdates.htm

But that link is wrong.. The real update to play Xbox games on Xbox 360 is here:
http://download.microsoft.com/download/d/1/8/d181ee58-de70-4484-936b-0e9161ccd6b2/BackCompat_11-2007.zip

You must have a Xbox 360 hard drive in order to apply this update and to the best of my knowledge it can only be applied by buring the file to a CD or by connecting to Xbox Live. You cannot apply with USB drive

Applying both update 9199 and the SystemUpdate_12-2007 will not do anything to a modded Xbox.

Media and Xbox 360

As best I can tell, there are two ways to play media on the Xbox 360.

Through Media Center
Video Library

Media Center

Media Center does not support playback of XVIDs. The updates that add XVID support do not apply to Media Center. Which makes Media Center pretty useless, for me anyway.

I believe the update that adds this is 2.0.6683.0.
http://en.wikipedia.org/wiki/Xbox_360_System_Software

Video Library

This is the big one. From here you can play XVIDs, connect to streaming servers (which I will cover later). But there is a big gotcha. In order to make it do all this great stuff you need the Optional Media Update.
http://marketplace.xbox.com/en-US/games/media/66acd000-77fe-1000-9115-d802fffe07df/?of=3

From what I have read, the Optional Media Update comes with some limitations:

Can only be downloaded from Xbox Live
It is signed to your console, if you take your HDD to your friends house and download it and bring it back to your house it will not work unless you connect to Xbox Live

If the update is installed correctly it will work while not being connected to Xbox Live.

I have read some sketchy tutorials that say you can hack it onto the Xbox 360 HDD but I doubt that work because they do not address the signing. I have not personally tried.

You will also need a Xbox 360 HDD to get this update. I believe all of this is part of Microsoft's plan to punish those with modded consoles.

If you do not have the update you will get the following error code:

51-coodf236

Streaming Media to your Xbox 360

If you want to play movies on your 360 you can either copy the files to/stick in a USB thumb drive or you can stream directly to the Xbox.

Two popular options exist for the streaming:

I have only used PS3 Media Server and it's great. Really good.

Here is a link all the video formats that Xbox 360 supports:
http://support.xbox.com/support/en/us/nxe/gamesandmedia/movies/videofaq/viewvideoplaybackfaq.aspx

Conclusions

Hopefully this info hels out somebody!

Getting started with X10

mike.clarke — Wed, 28 Oct 2009 15:27:00 GMT

Getting started with X10

If your anything like me you are always looking for some sort of IT project. Recently I started messing around with X10 and here is a summary of what I have learned.

X10 is home automation technology that uses your homes electric system to send/receive commands. X10 old but still the most popular home automation technology. It has retained its popularity mostly due to the fact the it is a open standard. But the technology definitely has its problems, like it isn't 100% reliable.

Here are some wiki links to X10 and X10's competitor technology:

X10
X10 is old and sometimes problematic but the most popular.
Insteon
Insteon backwards compatible with X10 but uses more a mesh technology. It is proprietary and more expensive than X10.
Z-Wave
Proprietary wireless home automation.
UPB
Similar to X10 but not nearly as popular.

X10 Pitfalls

X10 Summary

Simply put, X10 is made up of Controllers and Modules. Controllers broadcast signals to modules through your homes electric wiring and the modules change state when they receive a command addressed to them (ON/OFF). There is also a RF standard for X10 controllers, meaning that you can issue X10 commands using a RF remote control.

Signal Loss/Filters

X10 has a bad reputation for not working. The main reason it doesn't work is because there is noise on the line. You get noise on the line when you have something plugged causes interference, like a TV, Fridge, or Computer. In order to ensure reliable signal transmission you need filters on these devices.

Here are some links to some filters:
http://www.smarthome.com/1626/FilterLinc-Plug-In-X10-Noise-Filter/p.aspx
http://www.smarthome.com/1626-10/FilterLinc-10-Amp-Plug-In-X10-Noise-Filter/p.aspx
http://www.smarthome.com/4845ACF/15-Amp-Plug-In-Noise-Filter-AF120/p.aspx

Reaching all Modules/Phase Coupler

If your in North America (and most likely other countries) you might find it is impossible to control all of your modules from one controller, from one location. In short your 120V is divided into two phases and signals from controllers plugged into one phase cannot "jump" to the other phase.

You can read more about it here:
http://kbase.x10.com/wiki/Phase_Coupling

The easiest solution I have found in a Phase Coupler that plugs into your dryer outlet. Below is the link:
http://www.smarthome.com/4816A2/SignaLinc-Plug-In-Phase-Coupler/p.aspx

Computer Driven Home Automation

Controllers

If you get into X10 and get a collection of modules working the next step is of course to hook it up to your computer. The first thing you are going to need to do is connect your computer to your X10 system. There are several different components that can do this. Here is a short (most likely incomplete) list.

Internet searches return many more results. Most of the controllers that come back are USB Insteon controllers that are backwards compatible with X10.

Software

There are many, many different home automation packages. Here is a short list of some of the major names:

ActiveHome
ActiveHome Pro
Market leading in home automation software.
MisterHouse
Popular open source solution. Does everything but is very complicated.
X10Server
Open source client/server java app.

I personally use X10Server. It has the basic functionally I was looking for. Client/Server app and it allows you to schedule jobs in a ini file. And it is pretty simple to use. The only setback is that it is difficult to get running. It has very specific JVM requirements.

Conclusion

Home Automation is very time consuming, technically complicated, and somewhat expensive. Despite a lot of hard work X10 systems can still become completely unreliable if you plug a appliance that creates a lot of noise on the line. But if you enjoy messing around with technology and look at these types of projects as entertainment it can be pretty fun!

Hopefully you have found this guide useful.

IP Cam Streaming to Blackberry

mike.clarke — Tue, 22 Sep 2009 22:55:00 GMT

IP Cam Streaming to Blackberry

So you have a IP Cam and that's pretty neat. But now you want to be able to check in on it from your Blackberry. Well get ready, this is going to be crazy.

Before getting into it, it might be worth it to review the following:
(Terrible) Adventures with WVC54GC IP Cam
(Notice the word Terrible)

Review of Streaming Technology

Understand some of the following:

RTSP: This is a protocol for streaming media to clients. It works like HTTP over Port 554. The client makes a request and the Streaming Server responds.
RTP: Protocol used to deliver audio and video. RTSP goes over RTP.
3GP: This is a container format for the stream that the Streaming Server sends out. Most phones (like the Blackberry) support 3GP.
Darwin Streaming Server: DSS is a streaming server that supports 3GP. A streaming server is like a web server but for streams.
Windows Media Services: We aren't going to use this but I just included it to illustrate that there are different streaming servers.
VLC: Amazing media player among many other things.
SDP: This is file that VLC spits out to describe the outgoing stream to DSS.

Blackberry Support Streams

Review the following:
http://www.blackberry.com/btsc/search.do?cmd=displayKC&docType=kc&externalId=KB05482&sliceId=SAL_Public&dialogID=185006185&stateId=1%200%20185004605

According to that, all Blackberries support RTSP, 3GP and MPEG4.

Solution

The easiest solution is to buy a pricey IP cam that supports MPEG4 stream. Many do as viewing the cam on your phone is a attractive feature. But if you cheaped out (like me) you will need something else:

VLC is going to pull in the stream from the cam, transcode it to MPEG4, write out the transcoded stream using RTP to DSS (announce itself to DSS). The Blackberry is going to make a request to the streaming server (like a browser request) and boom! If everything goes well the DSS will serve up the live stream.

Alternatively, VLC has a web server (a streaming server) built into it. I couldn't find any documentation on that, but it may be possible to use VLC instead of DSS.

Install and Configure DSS for Windows

Download and install ActivePerl
Download and install DSS

DSS requires ActivePerl to run the web based admin console. ActivePerl can be downloaded from here:
http://www.activestate.com/activeperl/

Download DSS for Windows here:
http://static.macosforge.org/dss/downloads/

Here is a guide on setting up DSS. Get it installed and tested:
http://generally.wordpress.com/2007/08/07/how-to-setup-darwin-streaming-server-on-windows/

Install VLC for Windows

VLC for Windows can be downloaded here:
http://www.videolan.org/vlc/download-windows.html

Configure VLC to RTS to DSS

Review the following:
http://wiki.videolan.org/Documentation:Streaming_HowTo/Advanced_Streaming_Using_the_Command_Line

Start VLC in command line mode with all the parameters to create a sdp in the media folder of DSS.

This one line took me a very long time to figure out. I wanted to run this on Ubuntu but I could never get it to run stable. Oddly enough it works perfectly on Windows.

Replace http://192.168.0.152/img/video.asf with the URL with whatever stream comes out of your IP Cam.

C:\Program Files\VideoLAN\VLC\vlc.exe -I dummy http://192.168.0.152/img/video.asf ":sout=#transcode{vcodec=mp4v,width=80,height=60}:rtp{dst=127.0.0.1,sdp='file://C:/Program Files/Darwin Streaming Server/Movies/stream.sdp'}" --noaudio -v

Test the Deployment

Open up Quicktime and open up the URL rtsp://server/stream.sdp if it works, try it on your Blackberry!

Conclusion

Hope you found this guide useful.

Installing pkgutil and webmin on Solaris

mike.clarke — Thu, 27 Aug 2009 17:51:00 GMT

Installing pkgutil and webmin on Solaris

pkgutil

blastwave.org is a group of people who create Open Source software packages for the Solaris Operating System. Ensure you the right package for your Architecture.

They use to provide a tool called pkg-get to download Solaris packages, but that tool fell into legal troubles and has now been abandoned. The replacement tool is called pkgutil. It can be downloaded here:
http://www.blastwave.org/howto.html

It can be installed using the following:

# mkdir packages # cd packages/ # /usr/sfw/bin/wget http://blastwave.network.com/csw/pkgutil_sparc.pkg --13:10:04-- http://blastwave.network.com/csw/pkgutil_sparc.pkg => `pkgutil_sparc.pkg' Connecting to blastwave.network.com:80... connected! HTTP request sent, awaiting response... 200 OK Length: 347,136 [application/octet-stream] 0K .......... .......... .......... .......... .......... 14% @ 49.75 KB/s 50K .......... .......... .......... .......... .......... 29% @ 82.64 KB/s 100K .......... .......... .......... .......... .......... 44% @ 107.30 KB/s 150K .......... .......... .......... .......... .......... 58% @ 163.93 KB/s 200K .......... .......... .......... .......... .......... 73% @ 164.47 KB/s 250K .......... .......... .......... .......... .......... 88% @ 29.64 KB/s 300K .......... .......... .......... ......... 100% @ 62.40 KB/s 13:10:14 (67.83 KB/s) - `pkgutil_sparc.pkg' saved [347136/347136] # pkgadd -d ./pkgutil_sparc.pkg The following packages are available: 1 CSWpkgutil pkgutil - Installs Solaris packages easily (sparc) 1.6.1,REV=2009.06.15 Select package(s) you wish to process (or 'all' to process all packages). (default: all) [?,??,q]: Processing package instance <CSWpkgutil> from </packages/pkgutil_sparc.pkg> pkgutil - Installs Solaris packages easily (sparc) 1.6.1,REV=2009.06.15 Please see /opt/csw/share/doc/pkgutil/license for license information. ## Processing package information. ## Processing system information. ## Verifying disk space requirements. ## Checking for conflicts with packages already installed. ## Checking for setuid/setgid programs. This package contains scripts which will be executed with super-user permission during the process of installing this package. Do you want to continue with the installation of <CSWpkgutil> [y,n,?] y Installing pkgutil - Installs Solaris packages easily as <CSWpkgutil> ## Installing part 1 of 1. /etc/opt/csw/pkgutil.conf.CSW /opt/csw/bin/bldcat /opt/csw/bin/chkcat /opt/csw/bin/pkgutil /opt/csw/etc/pkgutil.conf.CSW /opt/csw/libexec/pkgutil/wget /opt/csw/share/doc/pkgutil/license /opt/csw/share/doc/pkgutil/readme /opt/csw/share/man/man1/bldcat.1 /opt/csw/share/man/man1/chkcat.1 /opt/csw/share/man/man1/pkgutil.1 /var/opt/csw/pkgutil/admin.CSW [ verifying class <none> ] ## Executing postinstall script. Installation of <CSWpkgutil> was successful.

If you add /opt/csw/bin path you can use pkgutil and all the software downloaded from blastwave.

webmin

webmin is great web management console for administering unix machines brought to you by the good people of http://www.webmin.com/. There is Solaris package of it that can be downloaded and installed. Ensure you the right package for your Architecture.

# mkdir packages # cd packages/ # /usr/sfw/bin/wget http://prdownloads.sourceforge.net/webadmin/webmin-1.480.pkg.gz pkgadd -d webmin-1.480.pkg WSwebmin ... Installation of <WSwebmin> was successful.

You can then access webmin web console using the following URL:
http://servername:10000

See http://www.webmin.com/solaris.html for more details.

Playing Worms Armageddon over the Internet

mike.clarke — Mon, 27 Jul 2009 02:14:00 GMT

Playing Worms Armageddon over the Internet

Worms Armageddon is a great game. Maybe you played in University over LAN and now you want to play it over the Internet. Inside Worms there are two multi player options: LAN and WormNET.

I have no idea what WormNET is, I've never been able to make it work. This guide is going to explain how to play a LAN game over the Internet.

LAN's and Worms

A LAN is a group of computers connected to each other that can talk to each other. LANs are great for gaming and a LAN is required for Worms multi player.

When you have the Internet in between a bunch of computers, the LAN model doesn't work anymore and neither does multi player worms.

So the answer is obvious, make a LAN over the Internet. This can be done through Virtual Private Networking (VPN). VPN is the technology that allows you to create LANs over the Internet.

Playing Worms Armageddon over the Internet

Hamachi is a free VPN service that works really well for gaming because it route's UDP packets, something that not all VPN does, like PPTP. So download and install Hamachi. You can download Hamachi here:
https://secure.logmein.com/products/hamachi/vpn.asp?lang=en

Once you've downloaded, installed, configured Hamachi it creates a additional NIC on your computer. This NIC is going to used in the VPN connection and is going to fool Worms into thinking it is on a LAN -- so you can player LAN games over the Internet.

So in Hamachi you can make password protected groups. These groups are your LAN, you can create a group and send the name and password to your friends and instand LAN, instant worms LAN party.

Multi Homed PC's and old Games

So with Hamachi installed you will have at least two NICs. Some old games don't function right with two NICs installed. They don't know which NIC to use. I think worms is one of these games. I'm not sure, but this will ensure that it will definately work.

Download this program called ForceBindIP. Get the ForceBindIP-1.2-Setup.exe. It can be downloaded here:
http://www.r1ch.net/stuff/forcebindip/

So ForceBindIP makes a program only see one NIC. We are going to use it force worms to only see one NIC.

So create a new Windows shortcut and in the location of the item type the following:

%windir%\system32\forcebindip.exe HAMACHI_IP "C:\MicroProse\Worms Armageddon\wa.exe"

But replace HAMACHI_IP with the IP address you got from Hamachi.

Update Worms

Worms v3.0 crashes for me when using my Hamachi shortcut. You need to get the latest patch to make it work. The latest version of worms is v3.6.29.0 Beta. The patch can be downloaded here:
http://wormsarmageddon.team17.com/main.html?page=supp&area=upda&file=15

Install it.

Worms!!!

Finally, get on Skype, connect up with Hamachi, start worms through the customer shortcut and let the sheep fly!

Running Mecury Mail as a Service

mike.clarke — Mon, 15 Jun 2009 18:24:00 GMT

Running Mecury Mail as a Service

Mercury mail is a great, free mail server that can be used for dozens of things. Unfortunately the free version does not run as a a service.

This post will explain how to run the free version of Mercury Mail as a service. This procedure could be extended to run other processes as services as well.

Creating a Customer Service

Review the following KB article:
http://support.microsoft.com/kb/137890

In short, you can use Instrsrv.exe to create your own services and srvany.exe to start executables. Instrsrv.exe and srvany.exe are included in the Windows Server 2003 Resource Kit Tools which can be downloaded here:
http://www.microsoft.com/downloads/details.aspx?FamilyID=9D467A69-57FF-4AE7-96EE-B18C4790CFFD&displaylang=en

Both instsrv.exe and srvany.exe have worked for me on x64 Server 2008

Creating the Mercury Mail Service

Create a folder called service in your MERCURY installation folder. Copy instsrv.exe and srvany.exe. Create to batch scripts, call one service.bat and the other remove.bat. You should have something that looks like this:

Add the following to service.bat (this batch file will create the service)

"C:\MERCURY\SERVICE\Instsrv.exe" "Mercury for Win32" "C:\MERCURY\SERVICE\Srvany.exe" pause

Add the following to remove.bat (this batch file remove/uninstall the service)

"C:\MERCURY\SERVICE\Instsrv.exe" "MERCURY32" REMOVE pause

Next run service.bat. This will create the service entry.

The final step is to setup the registry for srvany.exe. Create the following keys and values:

Finally start up the service.

Be careful not to start Mercury in the native app mode while the service is running, or you might make it angry. The best way to do it is to stop the service, launch Mercury, make your configuration changes, close Mercury, restart the service.

Conclusion

Hope you found this information useful!

Using Your Laptop As A Wireless Router

mike.clarke — Fri, 05 Jun 2009 01:09:00 GMT

Using Your Laptop As A Wireless Router

So for some reason you want to use your laptop as a wireless router.

This can be done using the Windows Internet Connection Sharing feature. This guide will cover how to set it up for Windows XP. For more information see:
http://support.microsoft.com/kb/306126
http://en.wikipedia.org/wiki/Internet_Connection_Sharing

If you review the above links you will see they don't mention how to use ICS to turn your laptop into a Wireless Access Point. So here how you do it.

You cannot use ICS if the network you are sharing has a IP range of 192.168.0.*

Step 1 - Create Your Wireless Network

Create your Ad Hoc wireless network

I personally like WPA because I find WEP keys very hard to remember.

In order to make this connection always available, ensure the following setting.

Step 2 - The Big Secret

In order for ICS to work with your wireless NIC you must set your wireless NIC with the following static settings:

Thank you so much to this post!
http://shepherdnick.wordpress.com/setting-up-a-pc-as-a-wireless-router-using-ics/

You will need to change this back to use any other wireless network with your laptop.

Step 3 - Enable ICS

On your physical network (the connection you want to share) enable ICS

Step 4 - Start Using Your Network

Join your wireless network the way you would join any other wireless network.

Unanswered Questions

I'm not sure if it is possible to not broadcast the SSID of your network. I was unable to locate any information on this. If anybody know please but it in the comments.

Conclusion

Hope you find this guide useful. You wouldn't believe how long it took me to figure this out.

32bit Windows Memory and Java

mike.clarke — Wed, 27 May 2009 20:56:00 GMT

32bit Windows Memory and Java

Understanding 32bit process limitations in Windows is complex. The following is a quick run down of basics, followed by a quick explanation of JRockit (and why it rocks!).

The Problem

Physical Address Extension

So as you can see from above, Physical Address Extension allows a 32bit OS to address up to 64GB of RAM. 32bit Windows can in fact have up to 64GB of RAM depending on the version. See the following to learn about the memory limits of each version of Windows and PAE:
http://msdn.microsoft.com/en-us/library/aa366778.aspx
http://www.microsoft.com/whdc/system/platform/server/PAE/pae_os.mspx
http://en.wikipedia.org/wiki/Physical_Address_Extension

In short, if you run 32bit Server 2003 Enterprise you can put up to 64 GB of RAM in your machine (provided your hardware has PAE). From a process perspective, PAE is unimportant.

Process Address Space (No /3GB switch)

Now we talking about how much memory a 32bit process can use in Windows. By default without modifications the theoretical limit is 2GB. In practice the limit is usually around 1.5GB.

This is because for a 32bit process Windows cuts the Process Address Space in half. Half goes to Kernal Memory (which a user process cannot use) and the other half goes to user mode (which you can use). It doesn't matter how much RAM is in the box, a 32bit process can only use 2GB of RAM. See below for more info:
http://msdn.microsoft.com/en-us/library/ms189334.aspx

Process Address Space (With /3GB switch)

When you use the /3GB in the boot.ini what this does is it decreases the Kernal mode memory and increases the user mode memory. But there is a catch, the program must be compiled/linked using the /LARGEADDRESSAWARE switch.

See the following under the Application Changes section:
http://www.microsoft.com/whdc/system/platform/server/PAE/PAEmem.mspx

Address Windowing Extensions

The final thing is Address Windowing Extensions. This when a program is written to use more memory that what is in the Process Address Space. I think Exchange, and SQL have this ability. Below is more information:
http://en.wikipedia.org/wiki/Address_Windowing_Extensions

Summary

Some hopefully that helps to focus what's going on with 32bit processes in Windows. If your into Java hopefully know your asking yourself is the JRE compiled with the /LARGEADDRESSAWARE switch?

Finding the answer to that has proven very difficult, but the answer is a disappointing no.

This is the best answer I could find. See:
http://download.boulder.ibm.com/ibmdl/pub/software/dw/jdk/diagnosis/dw3gbswitch3.pdf

This is the important part:

So what does the /3GB switch do for Java then?
The /3GB switch does enable the Java heap to grow a bit from a practical max of about 1.5 GB up to 1.7 GB or maybe even 1.8 GB. If you are already struggling to fit everything into your Java heap, the /3GB switch is not going to be of much help. The key advantage lies in the extra space available for threads (stack space), and native code (either for JIT’d code or application requirements).

Below is another explanation:
http://blogs.sun.com/moazam/entry/why_can_t_i_allocate

JRocket

JRocket is the Oracle version of the JRE. It is compiled with the /3GB switch. In addition it doesn't need a continues Process Address Space.

Usually when the Sun JRE/JDK asks Windows for a ton of memory, say like 1.5GB. What it really says to Windows is I need 1.5GB of continues memory. This can be a hard order to fill for the OS depending on what it is doing at the time. JRocket doesn't require the memory be continues which is great!

More information about JRocket can be found under the following links:
http://www.oracle.com/appserver/jrockit/index.html
http://www.oracle.com/technology/software/products/jrockit/index.html
http://web.archive.org/web/20080119012415/http://dev2dev.bea.com/blog/hstahl/archive/2005/12/how_to_get_almo.html

Just to illustrate that drastic difference you can get between the Sun JRE/JDK and JRocket below are some screen shots from a JBOSS server. The box has 5GB of RAM and the /3GB switch enabled. Using the standard JRE I can't start JBOSS with more than 1536MB of memory. Using JRocket I can start it with a whopping 2560MB of memory.

With Sun JDK

set JAVA_OPTS=%JAVA_OPTS% -Xms2560m -Xmx2560m

With JROCKET

set JAVA_OPTS=%JAVA_OPTS% -Xms2560m -Xmx2560m

Conclusion

JRocket kicks ass! Not only can it use more memory than the Sun JRE it's also optimized for speed!

Tomcat Authentication to Active Directory

mike.clarke — Fri, 22 May 2009 15:33:00 GMT

Tomcat Authentication to Active Directory

Love IIS and its easy Authentication Model (Anonymous, Basic, Integrated)? Me too! The following post will, to the best of my ability, explain how to make Tomcat to Basic and Integrated Authentication to Active Directory (AD). It actually isn't to bad but the information is all over the place.

In order to be successful at this you will need to have a good working knowledge of Tomcat. If you don't I suggest reading this:
http://mystyleit.com/blogs/mystyleit/archive/2008/12/05/adventures-with-tomcat-5-5.aspx

Tomcat LDAP Authentication to AD (aka Basic)

If you have read the official Tomcat documentation, hopefully you know that we need to create a JNDI Realm. In this example we are going to create a Global Realm, the means it is going in the Service area of the server.xml.

Here is the link to the official documentation:
http://tomcat.apache.org/tomcat-5.5-doc/realm-howto.html#JNDIRealm

If you have read the documentation you will see that JNDI Realms support to modes, Bind mode and Comparison mode. Based on everything I have read, Bind mode does NOT work with AD. Only Comparison mode modes does.

Creating the authentication is a two step process. The steps are:

We must create a global JDNI Realm
Create a security constraint in the web.xml (equivalent checking off the Basic Authentication check box in IIS)

So lets have a look at a Comparison mode JNDI Realm. This would go in the server.xml:

<Realm className="org.apache.catalina.realm.JNDIRealm"  connectionURL="ldap://dc1.domain.local:389"  connectionName="tomcat@DOMAIN" connectionPassword="AReallyLongPassword"  userBase="OU=users,OU=mydomain,DC=DOMAIN,DC=local" userSearch="(sAMAccountName={0})" userSubtree="true" userRoleName="memberOf"  roleBase="OU=groups,OU=mydomain,DC=DOMAIN,DC=local" roleSubtree="true" roleName="cn" roleSearch="(member={0})"  debug="99" />

So lets have a look security constraint. This goes in your applications web.xml:

<security-constraint> <web-resource-collection> <web-resource-name>Entire Application</web-resource-name> <url-pattern>/*</url-pattern> </web-resource-collection>  <auth-constraint> <role-name>Tomcat Users</role-name> </auth-constraint> </security-constraint>  <login-config> <auth-method>BASIC</auth-method> <realm-name>JNDIRealm</realm-name> </login-config>

Below is a link to a great MSDN web cast about setting up the same thing:
http://blogs.msdn.com/alextch/archive/2007/06/25/configuring-tomcat-to-authenticate-against-active-directory.aspx

Tomcat NTLM Authentication to AD (aka Integrated)

Next up is setting up Tomcat to use NTLM authentication (aka Integrated). This one is more complicated.

NTLM

NTLM is complicated stuff. In short there are two versions of NTLM, NTLMv1 and NTLMv2. NTLMv2 is the good stuff. NTLMv1 is the bad stuff. If you would like to know more I recommend reading the following:
http://en.wikipedia.org/wiki/NTLM
http://technet.microsoft.com/en-us/magazine/2006.08.securitywatch.aspx

JCIFS

JCIFS is a open source client library that can do all sorts of things, including to NTLM Authentication for a web app. See the following:
http://jcifs.samba.org/
http://jcifs.samba.org/src/docs/ntlmhttpauth.html

Deploying JCIFS

There are many deployment scenarios for JCIFS. The deployment I'm covering getting JCIFS to Authenticate against a Network Share. In this approach you create a secure network share, only the users who have access to the network share can access the web site.

The installation has several steps:

Create a secured share and secure using NTFS permissions
Download the jcifs.jar and copy to the lib folder of your web app
Create the JCIFS filter in you web app's web.xml file

This guide is not going to cover how to create a network share. So look elsewhere.

Let's have a look at the JCIFS filter we need to create in the web.xml:

<filter> <filter-name>NtlmHttpFilter</filter-name> <filter-class>jcifs.http.NtlmHttpFilter</filter-class> <init-param> <param-name>jcifs.http.domainController</param-name> <param-value>192.168.0.30</param-value> </init-param> <init-param> <param-name>jcifs.smb.client.domain</param-name> <param-value>CITY</param-value> </init-param>  <init-param> <param-name>jcifs.smb.client.logonShare</param-name> <param-value>JcifsAcl$</param-value> </init-param> </filter>  <filter-mapping> <filter-name>NtlmHttpFilter</filter-name> <url-pattern>/*</url-pattern> </filter-mapping>

For more information see:
http://jcifs.samba.org/src/docs/ntlmhttpauth.html#install

The above configuration works on my Server 2008 native domain with no change to the Domain Security policy.

Conclusion

When I went to set all this up myself the first time I had a really hard time finding examples of configurations. Hopefully these examples are helpful to you.

(Terrible) Adventures with WVC54GC IP Cam

mike.clarke — Mon, 11 May 2009 18:30:00 GMT

(Terrible) Adventures with WVC54GC IP Cam

I was looking for a new, relatively inexpensive, gadget for my house. I was looking at the flyer for Canada Computers and there was a deal on a Linksys WVC54GC IP Cam. So I figured OK.

Anybody who has bought a WVC54GC knows that this IP Camera has a complicated setup and not to many features. I learned this the hard way. The positive thing about this camera is that, if your crafty, you can make it do everything with a lot of work. What following is the things I managed to make mine do.

Streams Coming out of the Cam

The cam has three streams coming out of the cam:

http://[camera address]/img/video.asf [ASF Stream]
http://[camera address]/img/mjpeg.jpg [Motion JPEG]
http://[camera address]/img/mjpeg.cgi [Motion JPEG]

Opening one of the Steams

The cam is designed to be used either through Internet Explore (via ActiveX control) or using the Linksys IP Cam think client, available for download here:
http://www.linksysbycisco.com/CA/en/support/WVC54GC/download
The Setup Wizard is actually the Think Client

However the above mentioned streams can also be directly opened in VLC. VLC is a cross-platform media player and streaming server.
http://www.videolan.org/vlc/

Capturing a Still Image from the IP Cam

So as complained about by many, the Cam cannot capture a still jpg. Can't do it. The WVC54GCA can but not this guy. Plus it needs a ActiveX control to view anything so, checking in on the cam from any browser just isn't possible.

Solution

So here starts the terrible adventure. There are ton of open source encoding tools that can literally do everything in the right hands. As with most open source projects, they are not that polished and more geared for Linux users. One of them is mplayer.exe. The mplayer project can be found here:
http://www.mplayerhq.hu/

As you may have noticed, only the source code is available. Thankfully, somebody has compiled it for Windows and posted it here and lots of other places:
http://majorgeeks.com/MPlayer_d5663.html/

So it is possible to grab a frame from the stream as a jpeg using the following mplayer.exe command

mplayer.exe http://[camera address]/img/video.asf -frames 1 -vo jpeg:quality=100:maxfiles=1

I got this info from this guy:
http://www.infohit.net/blog/post/using-linksys-wvc54gc-webcam-with-linux.html/

I took this caputing a jpg from the cam one step futher. I wrote a .NET/C# app to execute a batch script (that contains the above command). This bascially creates a event driven web site that can display a still from the cam. This is a very rough app. This app requires framework 2.0.

This small (and terrible) little app can be downloaded here:
http://mystyleit.com/downloads\blogs\TerribleAdventureswithWVC54GCIPCam/poll.zip

Capturing a Video from the stream

The next thing I wanted to do with this cam was to capture a video at a specific time. This can be done through the Thick Client but I really don't like the included scheduler. I'd prefer to do it with a batch script stuffed into the Windows Scheduler.

Solution

A command line solution does exist using a open source project called ffmpeg. The ffmpeg project is located here:

http://www.ffmpeg.org/

Like mplayer this project doesn't come compiled (and not compiled for windows for sure). Luckily somebody else has compiled it and this is located here:
http://www.videohelp.com/tools/ffmpeg

ffmpeg.exe -i http://[camera address]/img/video.asf -vcodec copy -b 900k output.asf

Most of this cam from this post:
http://www.infohit.net/blog/post/using-linksys-wvc54gc-webcam-with-linux.html/

A asf file should be able to play in Windows Media Player without issue. I suppose you could re-encode the file to a mpeg or something else depending on your requirements.

Transcoding to Flash

This is where things go terrible. So what if you want to see the stream in a browser but don't want to install the ActiveX control? Well you can do on the fly transcoding to a different stream type, like flash

According to this this you should be able to open the asf stream in a embedded media player, but I was never able to make this work.

I believe this CAN be done through VLC but I have not been able to figure out how, or been able to find any posts about how to do this whatsoever. If anybody knows how please comment on this post or send me a message here. See link below for more information about VLC:
http://wiki.videolan.org/Flash_Video

If however you are game, this can be done through ffserver. To the best of my knowledge there is no port of ffserver to Windows. I am unsure if it can be compiled for Windows through Cygwin. I tried but get the same errors as everyone else. Ever post I found did not have a solution. If you are looking for a guide to proceed this is the best I have found:
http://www.infohit.net/blog/post/motion-capture-using-the-wvc54gc-with-linux.html
http://ubuntuforums.org/archive/index.php/t-665607.html

Here is a excellent guide on how to install ffmpeg (and ffserver) on a Ubuntu box:
http://ubuntuforums.org/showpost.php?p=6963607&postcount=360

If you can follow all that, if you add the following to your ffserver.conf file, you'll get a swf stream out that you can open in a browser.

<Stream stream1.swf> Feed feed1.ffm Format swf VideoCodec flv VideoFrameRate 2 VideoBufferSize 80000 VideoBitRate 100 VideoQMin 1 VideoQMax 5 VideoSize 352x288 PreRoll 0 Noaudio </Stream>

If you put something like the following in a page, you should be able to open the swf stream:

See the following for more information:
http://ubuntuforums.org/showthread.php?t=665607

Again, I think it should be possible to do the transcode through VLC but I have not be able to figure it out. Hopefully somebody can post some tips about this.

Running FFMPEG using a Script

The FFMPEG process needs a terminal to run. In addition it seems to crash every 5 to 6 hours. Ideally you would have a cron job to monitor the process but running FFMPEG in a script can be tricky. It can be done using the screen command.

Here is my cron job script which I have scheduled to run every hour. I kill FFMPEG because I found that sometimes the process was still running but not working. This way I can ensure that I will always get a stream.

ffserver_pid=$(ps -e | grep ffserver | cut -c 1-6) ffmpeg_pid=$(ps -e | grep ffmpeg | cut -c 1-6) screen_pid=$(ps -e | grep screen | cut -c 1-6) if [ "$ffserver_pid" == "" ]; then echo "ffserver is not running..." echo "restarting ffserver." /usr/local/bin/ffserver fi if [ "$ffmpeg_pid" != "" ]; then echo "ffmpeg is running..." echo "restarting the process..." kill $ffmpeg_pid screen -d -m /usr/local/bin/ffmpeg -an -i http://[CAM ADDRESS]/img/video.asf http://localhost:8090/feed1.ffm else echo "fmpeg is not running..." echo "starting ffmpeg." screen -d -m /usr/local/bin/ffmpeg -an -i http://[CAM ADDRESS]/img/video.asf http://localhost:8090/feed1.ffm fi

Conclusion

Streaming media is complex. I much prefer web and app servers!

Adventures with Tomcat 5.5

mike.clarke — Fri, 05 Dec 2008 22:03:00 GMT

Adventures with Tomcat 5.5

Tomcat has one main configuration files: server.xml. There is also another very misunderstood file web.xml that I will talk about too.

server.xml contains all the parameters of the Tomcat instance. The file has the following "big parts":

NAME	FUNCTION	CAN CONTAIN
Server	Represents Tomcat itself	Service, Listener, GlobalNamingResources
Service	Groups Connectors that share an Engine.	One or more Connectors followed by one Engine, Listener
Engine	Handles all requests.	Host, Realm, Valve, Listener, Cluster
Host	One "virtual host".	Alias, Context, Realm, Valve, Listener, Cluster

Context	Configures one "web application" within a host.	Loader, Manager, Realm, Resources, ResourceLink, WatchedResource, Environment, Value, Listener

Now let's talk about the most important/common questions about the file, Connectors and Realms (ie Authentication).

Minimal server.xml File

You should use the default server.xml file, this is only presented in order to get familiar with it's layout.

Connectors

Tomcat has this idea of Connects, each Connector allows Tomcat to speak a different language, HTTP, or AJP, or HTTPS. The default Connectors are HTTP (on port 8080) and AJP (port 8009). AJP is what IIS or Apache talk to Tomcat on (when you have a web server sitting in front of Tomcat).

One important trick with the AJP connector is to turn off Tomcat Authentication when you have another web server sitting in front of Tomcat performing the authentication. To do this, add the following to your AJP connector entry:

SSL Connector

In order to have an SSL connector, you need to have a keystore that contains a SSL certificate. Here is a link to how to create a keystore, and other keystore tasks (import, export, etc):
http://java.sun.com/j2se/1.3/docs/tooldocs/win32/keytool.html

Here is a link to a GUI interface to the keystore, requires Java 6 and very handy:
http://yellowcat1.free.fr/

So (hopefully) you have a keystore with a key inside (the default self-signed key or something else). Now you need to add the SSL connector. It should go with the other connectors in the Service section and looks something like this:

In order to require SSL on a specific site, you need to configure a security constrant for that app. You can do this by editing that app's web.xml file (see below). The constrant looks like this: ><user-data-constraint> <transport-guarantee>CONFIDENTIAL</transport-guarantee> </user-data-constraint>

Deploying Web Appls and the web.xml File

Hopefully the above explains the server.xml enough to grasp what Tomcat is doing. The important thing to realize is that different sections of the file have differents scopes. The futher down you go, the narrower the scope.

Tomcat, like Apache, has this idea of a Virtual Host. Basically it allows you to run serveral different sites on the same port (either 80 or 8080). This post isn't going to cover all that. Instead, just realize that localhost is the default host.

In order to deploy a web app that isn't a WAR file (beyond the scope of this document) you need to create a context for it. You can either create the context in the host section in the server.xml (which is bad because it makes the file more complicated) or you can create a seperate xml file in the host folder, under the Engine.

The default host folder, (under the default engine) is of course:
\Tomcat 5.5\conf\Catalina\localhost

You should name the xml file will of the directory in the URL.
If I created \Tomcat 5.5\conf\Catalina\localhost\myapp.xml the URL for this application http://site/myapp

A very simple context looks something this:

For more about contexts see the following:
http://tomcat.apache.org/tomcat-5.5-doc/config/context.html

What about the web.xml?

Every web application is going to have a WEB-INF folder. In the above example, this would be C:\myapp\WEB-INF. Inside the WEB-INF folder is a file, web.xml.

The web.xml file in the WEB-INF folder of your web app is used to configure that web app. That is it's scope ONLY.

The web.xml file in Tomcat 5.5\conf is the global web.xml file. It is processed before it reads the WEB-INF\web.xml file in the app's docBase. You can override what is in the Tomcat 5.5\conf\web.xml by setting knew values in the WEB-INF\web.xml.

So (hopefully you have guessed this) you shouldn't mess about with the Tomcat 5.5\conf\web.xml file. It's way better to mess around with the web.xml in WEB-INF\

Realms (Tomcat Authentication)

Unlike IIS, Tomcat has many, different (and complicated) authentication methods. They call them Realms.

The steps for securing a web app is first to have a Realm setup in the server.xml file. The second is to create a security constrant in the app's web.xml file.

Going over how to create all the security constraint's is what Google is great for. The important thing to know about Realms are:

You make them in the server.xml file
Where you put them affects their scope (ie which applications can use the Realm)
When in doubt, put them under the Engine, that way every app run by that engine can use them.

Here is a link that covers all of Tomcat's Realms:
http://tomcat.apache.org/tomcat-5.5-doc/realm-howto.html

Conclusion

I love IIS. Unfortunately it doesn't run JSP and Servlets so you need an Application Server.

Adventures with x64 IIS

mike.clarke — Thu, 13 Nov 2008 22:35:00 GMT

Adventures with x64 IIS

Recently I had to deploy 32 bit application on a Windows x64 bit OS. This post raps up some of the things I found along the way.

ISAPI Filters

The application supports an ISAPI filter, but the filter is only 32 bit, therefore it doesn't work without modification. If you google this subject you might find this Microsoft KB article:

http://www.microsoft.com/technet/prodtechnol/WindowsServer2003/Library/IIS/405f5bb5-87a3-43d2-8138-54b75db73aa1.mspx?mfr=true

This is only a very small part of the entire story. In order make x64 IIS use x32 ISAPI DLLs you need to do the following:

Uninstall ASP.NET x64 (As if you switch x32 this will cause IIS to crash (ie: Service Unavailable Error)
To uninstall ASP.NET x64 open a command line window and navigate to:
C:\WINDOWS\microsoft.net\Framework64\v2.0.50727
Then execute the following command (this will uninstall ASP.NET x64):
aspnet_regiis -u
Now we will execute Microsoft's command, but you need to make sure you define the location of adsutil.vbs. If you haven't moved/deleted the file, the command should be:
cscript c:\inetpub\AdminScripts\adsutil.vbs set w3svc/AppPools/Enable32BitAppOnWin64 1
Finally, we will install ASP.NET x32. To install ASP.NET 32 bit open a command line window and navigate to:
C:\WINDOWS\microsoft.net\Framework\v2.0.50727
Then execute the following command:
aspnet_regiis -i

Okay, that was the first hurdle I had to get over.

Apache Tomcat Connector

So the application was deploying also has an Application portion part to it. Therefore I needed to make IIS play nice with Tomcat.

Here are the useful websites:
http://tomcat.apache.org/connectors-doc/
http://tomcat.apache.org/connectors-doc/webserver_howto/iis.html

You will need to use the x32 version of isapi_redirect.dll as we just switched IIS to x32 bit

You will also need to create the registry entries under Wow6432Node.

In short, the connector is controlled by two files, workers.properties and uriworkermap.properties.

workers.properties controlls the worker process. I have never needed anything more then the default/sample configuration:

# workers.properties.minimal -
#
# This file provides minimal jk configuration properties needed to
# connect to Tomcat.
#
# The workers that jk should create and work with
#

worker.list=wlb,jkstatus

#
# Defining a worker named ajp13w and of type ajp13
# Note that the name and the type do not have to match.
#
worker.ajp13w.type=ajp13
worker.ajp13w.host=localhost
worker.ajp13w.port=8009

#
# Defining a load balancer
#

worker.wlb.type=lb
worker.wlb.balance_workers=ajp13w

#
# Define status worker
#

worker.jkstatus.type=status

The other file, uriworkermap.properties, controlls the mappings, ie what things get forwarded to tomcat. Entries look like this:

/directory/*.do=wlb

In this case, everything in http://site/directory ending in do would be forwarded to tomcat.

Performance

My site in running under Hyper-V (which is really awesome, blows VMware Server out of Water). I'm not sure if it's because Java doesn't play nice with VMs, or something else. But every time Tomcat loaded my site, the CPU went crazy! I tried different versions on Java, and it still went crazy. The fix was NOT to install the native DLL.

URL Redirection

Finally, I needed to make a URL redirector. Nothing fancy, but here it is: <% Response.Redirect "http://site/isapi.dll" %>

In order for it to work, you will need to enable the Active Server Pages Web Service Extension.

Conclusion

Hope this information helps somebody fix something.

References

http://geekswithblogs.net/MattRobertsBlog/Default.aspx

MSCS Roundup

mike.clarke — Mon, 10 Nov 2008 21:23:00 GMT

MSCS Roundup

This article links to all the useful resources I have found that relate to Microsoft 2003 Failover Clustering.

MSCS Terminology

This article defines MSCS terminology:
http://technet.microsoft.com/en-us/library/cc757640.aspx

Server 2003 Cluster Models

There are 3 Cluster Models with Windows Server 2003.

Model 1: Single Node Cluster

In short, this Cluster Model would most likely only be used for testing.

Model 2: Single Quorum Cluster

Standard, Active/Failover (or Active/Active) Cluster.
What most people thing about when they think of MSCS Clustering.

Model 3: Majority Node Set

Not common
Most likely used only by software designed for this type of cluster.

Model 4: [NEW] Majority Node Set (with a Witness)

See this KB for more information:
http://support.microsoft.com/kb/921181

Additional Information

This article explains the models in greater detail:
http://technet.microsoft.com/en-us/library/cc739522.aspx

Server 2008 Cluster Models

Cluster Models in Server 2008 have changed from Server 2003. A full discussion is beyond the scope of this post, but here is a link that discusses them further:
http://technet.microsoft.com/en-us/magazine/cc672627.aspx

Creating a UNC Alias for a Windows 2003 Server

mike.clarke — Fri, 31 Oct 2008 21:04:00 GMT

Creating a UNC Alias for a Windows 2003 Server

There are many reasons my you make want to make a CNAME/alias for a Windows 2003 File Server. These might include ensuring old shortcuts work after a server migration, you might have an application with UNC paths embedded that are very difficult, or impossible to change.

The Problem

Assuming that you have your DNS CNAME created properly, when you try to open the UNC with you'll be asked to Authenticate, (which shouldn't happen)

And even when you try to login, it doesn't work.

The Fix

In order to allow a Windows File sever respond to a alias you need to disable some security by way of editing the registry. These keys need to be created (or editted):

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\lanmanserver\parameters]
"DisableStrictNameChecking"=DWORD:00000001

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Lsa]
"DisableLoopbackCheck"=DWORD:00000001

More Information

Here are the links to the offical KB articles.

Enabling XDMCP for SLES 9

mike.clarke — Thu, 09 Oct 2008 15:03:00 GMT

Enabling XDMCP for SLES 9

Open /etc/sysconfig/displaymanager
Change DISPLAYMANAGER_REMOTE_ACCESS="yes"

Open the files /etc/opt/kde3/share/config/kdm/kdmrc and /etc/opt/kde3/share/config/kdm/kdmrc.SuSEconfig Find the XDMCP section and set Enable to true
[Xdmcp]
Enable=true
Willing=/etc/X11/xdm/Xwilling
Xaccess=/etc/X11/xdm/Xaccess

Open /etc/X11/xdm/Xaccess
Make sure there is no # to the left of the * in the following host line
* #any host can get a login window

Open /etc/X11/xdm/xdm_config
Change the DisplayManager.requestPort line to
DisplayManager.requestPort: 177

SLES 9 SP1 and Oracle 10g in VMware

mike.clarke — Wed, 08 Oct 2008 19:12:00 GMT

SLES 9 SP1 and Oracle 10g in VMware

This post will briefly document how to create a test environment running SLES v9 SP1 with Oracle v10g Release 2 [10.2.0.1.0] in VMware.

Installing SLES

Download the SLES media from Novell, here are the links:
http://download.novell.com/Download?buildid=1HYSkGYIYow~
http://download.novell.com/protected/Summary.jsp?buildid=IaBIrEFrE9Q~
http://wiki.novell.com/index.php/SLES_9_Download

You will need to register for a account

Create a new VM with at least 1GB of memory and boot using the Service Pack media, following the installtion guide.

Be sure to modify the default partition layout and ensure that they swap partition is double the size of the memory.
For example 1GB og physical memory would need a2GB swap partition. This is required by Oracle and it is NOT the default.

You can check what versio of SLES you have installed by executing the following command:

SPident

Installing VMware Tools

Next you need to install the VMware tools. Tell VMware to install the tools, them copy the install from /media/cdrom to /tmp/vm
Then cancel the installation of vmware tools.

Execute the following as root

mkdir /tmp/vm
cp /media/cdrom/*.tar.gz /tmp/vm/

Now we will actually install the tools. Navigate to /tmp/vm/ and unpack the installer. Then run the install script and accept all default values. I recommend setting the default login screen to 1074x768.

cd /tmp/vm/
gunzip *.gz
tar xvf *.tar
cd vmware-tools-distrib/ ./vmware-install.pl

Now reboot the system.

reboot

Preparing the System for Oracle

Hostname

Be sure to change the hostname and add it to the /etc/hosts file.

Required Packages

In order to install Oracle you will need to install the following pacakges

gcc
gcc-c++
glibc
libaio
libaio-devel
make
openmotif
openmotif-libs

These packages can be installed by executing the following commands:

yast -i gcc
yast -i gcc-c++
yast -i glibc
yast -i libaio
yast -i libaio-devel
yast -i make
yast -i openmotif
yast -i openmotif-libs

Kernal Parameters

Enable kernal parameters and add the following to /etc/sysctl.conf

chkconfig boot.sysctl on

Add these kernal parameters to /etc/sysctl.conf

kernel.shmmax=2147483648
kernel.sem=250 32000 100 128
net.core.rmem_default = 262144
net.core.rmem_max = 262144
net.core.wmem_default = 262144
net.core.wmem_max = 262144
net.ipv4.ip_local_port_range = 1024 65000

Reboot the system.

Creating the Oracle User Account

Execute the following commands as root to create the oracle user

groupadd oinstall
groupadd dba
useradd -g oinstall -G dba oracle
mkdir /home/oracle
usermod -d /home/oracle oracle
usermod -s /bin/bash oracle
chown -R oracle:oinstall /home/oracle
chmod -R 775 /home/oracle
passwd oracle

Next login as the Oracle user.
Get a terminal sesion going and create these files:

.bashrc
.profile

Add this to .bashrc:

. /home/oracle/.profile

This will make sure your .profile is executed every time you start a new terminal session

Add this to .profile

umask 022

ORACLE_BASE=/home/oracle; export ORACLE_BASE

ORACLE_SID=orcl; export ORACLE_SID
ORACLE_HOME=/home/oracle/product/10.2.0/Db_1; export ORACLE_HOME

PATH=$PATH:/usr/local/bin:/usr/sbin/:$ORACLE_HOME/bin; export PATH

Conclusion

You are now ready to install Oracle using the Oracle Universal Installer. Good Luck!

References

http://www.puschitz.com/InstallingOracle10gOnSUSE.shtml

Adventures with Solaris 10 SPARC-64

mike.clarke — Mon, 25 Aug 2008 20:11:00 GMT

Adventures with Solaris 10 SPARC-64

This article will cover how to create a Solaris 10 sparse-root Zone, how to install Oracle v10g Release 2 Apache HTTP Server v2.0.63 into the sparse-root Zone.

Solaris Zones Overview

A zone is a virtual operating system abstraction that provides a protected environment in which applications run. The applications are protected from each other to provide software fault isolation. To ease the labor of managing multiple applications and their environments, they co-exist within one operating system instance, and are usually managed as one entity.

In other words, a zone faux kernal and file system. In practice, they are very similar to Virtual Machines.

Zone Types

Global Zone

The actual host OS. All other zones run in the Global Zone

sparse-root Zone

Sparse-root zones optimize physical memory and disk space usage by sharing some directories, like /usr and /lib. Sparse-root zones have their own private file areas for directories like /etc and /var.

In other words, a sparce-root zone shares mounts most of the Global Zone's file system as Read-Only, makeing the zone much smaller than a Whole-root zone.

whole-root Zone

Whole-root zones increase configuration flexibility but increase resource usage. They do not use shared file systems for /usr, /lib, and a few others.

In other words, a whole-root zone is a copy of the whole root.

See the following for more information:
http://opensolaris.org/os/community/zones/faq/#basic_zone

Preparing Global Zone for Oracle v10g Release 2

Before installing Oracle 10g you first need to ensure that

All the required packages are installed
The kernal parameters have been configured

Packages

According to Oracle® Database Quick Installation Guide the following packages are required to install Oracle 10g

SUNWarc
SUNWbtool
SUNWhea
SUNWlibm
SUNWlibms
SUNWsprot
SUNWtoo
SUNWi1of
SUNWi1cs
SUNWi15cs
SUNWxwfnt
SUNWsprox

SUNWsprox can be disregarded, see http://forums.sun.com/thread.jspa?messageID=9329186

To verify that all the required packages are installed, run the following command:

pkginfo -i SUNWarc SUNWbtool SUNWhea SUNWlibm SUNWlibms SUNWsprot SUNWsprox SUNWtoo SUNWi1of SUNWi1cs SUNWi15cs SUNWxwfnt

SUNWi1cs and SUNWi15cs are not included in the Full Deployment of Solaris 10 and will need to be installed. There are included in the Solaris install media and can be installed using the pkgadd command:

pkgadd -d /cdrom/cdrom0/s0/Solaris_10/Product SUNWi1cs SUNWi15cs

Kernal Parameters

In previous versions of Solaris, kernel parameters were amended by adding entries to the "/etc/system" file, followed by a system reboot.
...The Oracle installer recognizes kernel parameters set using this method, but it is now deprecated in favour of resource control projects, explained below.

Make a copy of /etc/system (/etc/system.ORG) and add the following to end of the file:

set noexec_user_stack=1
set semsys:seminfo_semmni=100
set semsys:seminfo_semmns=1024
set semsys:seminfo_semmsl=256
set semsys:seminfo_semvmx=32767
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=10

Reboot the server

See the following for more information:
http://www.oracle-base.com/articles/10g/OracleDB10gR2InstallationOnSolaris10.php

Apache and GCC

In order to compile Apache HTTP Server a C compiler needs to be installed on the system. GCC is not included with Solaris and needs to be installed. Enter the following commands to install the needed packages

mkdir /tmp/gcc
cd /tmp/gcc
/usr/sfw/bin/wget/wget ftp://ftp.sunfreeware.com/pub/freeware/sparc/10/libiconv-1.9.2-sol10-sparc-local.gz
/usr/sfw/bin/wget/wget ftp://ftp.sunfreeware.com/pub/freeware/sparc/10/gcc-3.3.2-sol10-sparc-local.gz
gunzip gcc-3.3.2-sol10-sparc-local.gz
gunzip libiconv-1.9.2-sol10-sparc-local.gz
pkgadd -d libiconv-1.9.2-sol10-sparc-local
pkgadd -d gcc-3.3.2-sol10-sparc-local

You may wish to download a more recent GCC packages

See the following for more information:
http://frankmash.blogspot.com/2006/08/installing-gcc-compiler-on-solaris-10.html

Creating the sparse-root Zone

Preparing to Create Zone

Note the name of the physical adapter the zone is going to share.

# ifconfig -a
lo0: flags=2001000849 mtu 8232 index 1
    inet 127.0.0.1 netmask ff000000
hme0: flags=1000843 mtu 1500 index 2
    inet 192.168.132.36 netmask fffffe00 broadcast 192.168.133.255
    ether 8:0:20:d9:3a:a4

In the above example, you can see the adapter is named hme0 Now we will list all the running zones and create directory for the new zone

zonename
zoneadm list -iv
mkdir /export/home/myzone
chmod 700 /export/home/myzone

Creating and Installing the Zone:

zonecfg -z myzone
create
set zonepath=/export/home/myzone
set autoboot=true
add net
set physical=hme0
set address=XXX.XXX.XXX.XXX
end
verify
commit
exit

zoneadm -z myzone verify
zoneadm -z myzone install
zoneadm list -iv
zoneadm -z myzone boot

Use the followng command to connect the zone's console and complete the install:

zlogin -C myzone

Choose option 13 when prompted for console type

See the following for more information:
http://www.solarisinternals.com/wiki/index.php/Zones
http://os.miamano.eu/node/2

root Access

See my previous post for more details:
http://mystyleit.com/blogs/mystyleit/archive/2008/05/21/getting-started-with-solaris-10.aspx

Here is a very good guide about setting up a fresh Solaris deployment:
http://www.blastwave.org/howto.html

Creating Required Operating System Groups and User

According to Oracle® Database Quick Installation Guide the following groups and user must be created.

Groups

oinstall - The Oracle Inventory group
dba - The OSDBA group

User

oracle - The Oracle software owner

The following commands will create the needed groups, user (with bash shell and /export/home/oracle as home directory)

/usr/sbin/groupadd oinstall
/usr/sbin/groupadd dba
/usr/sbin/useradd -g oinstall -G dba oracle
mkdir /export/home/oracle
chmod -R 744 /export/home/oracle
chown -R oracle /export/home/oracle
usermod -d /export/home/oracle oracle
usermod -s /usr/bin/bash oracle
passwd oracle

If you plan to install oracle in the orcacle user's home directory, execute the following:

chown -R oracle:oinstall /export/home/oracle
chmod -R 775 /export/home/oracle

Configuring the oracle User's Environment

SSH in using putty as oracle and execute the following commands:

touch .profile touch .bashrc chmod 744 .profile chmod 744 .bashrc

Add the following line to .bashrc (gnome-terminal doesn't use .profile)

. ~/.profile

See the following for more information:
http://telin.ugent.be/~slippens/drupal/?q=bashrc_and_others

Add the following lines to .profile

umask 022

ORACLE_BASE=/export/home/oracle; export ORACLE_BASE

ORACLE_SID=orcl; export ORACLE_SID
ORACLE_HOME=/export/home/oracle/product/10.2.0/Db_1; export ORACLE_HOME

PATH=$PATH:/usr/openwin/bin:/usr/css/bin:/usr/sfw/bin:/usr/local/bin:$ORACLE_HOME/bin
export PATH

NOTE:During a normal Oracle installtion, /usr/local/bin/oraenv is created. Because we are installing Oracle into sparse-root zone /usr/local/bin is READ ONLY and therefore oraenv is NOT created. Therefore you need to manually set ORACLE_HOME and ORACLE_SID in accounts who rely on oraenv.

Running the Universal Installer

In order to run the Oracle Universal Installer, you need to be running X. If you are connecting to your SPARC-64 Solaris box from a Windows client I recommend you use PuTTY and Xming. They can be downloaded here:
http://sourceforge.net/projects/xming
http://www.putty.org/

Both of these are free and easy to install.

The normal way to X into your Solaris box is SSH in and redirect the Display Environment Variable. PuTTY can do this automatically. (Leave X Display Location blamk for local host

Once connected (or before, it doesn't matter) you need to start the X server on your Windows XP client. To do this start Xming (not XLaunch). This will start the X server. Notice the system tray icon.

Now if you have /usr/openwin/bin in your PATH variable, you should be able to start xterm.

This is how the Oracle Oracle® Database Quick Installation Guide tells you to run the installer...
This has NEVER worked for me. What happends is that the Universal Installer will seem to run fine but will NEVER be able to create the database. After accepting all your options about what database to create, it will ask you to confirm, then after that NOTHING. I think what happends is that it fails to create the Creating Database process window and therefore never creates the database.

The solution to my problem was to install Oracle using XDMCP, which loosely speaking is like RDP for Solaris.
To use XDMCP, start XLaunch and do the following:

Now if you launch gnome-terminal from the Start Menu, (recall gnome-terminal will use .bashrc not .profile (that is why we made .bashrc call .profile)) you can run the Universal Installer and everything should work as it should.

Starting Oracle after a Reboot

In order to start the Oracle processes after a reboot of the OS, login as the oracle user and execute the following commands:

sqlplus /nolog

connect / as sysdba
startup
show parameter db_name
quit

lsnrctl start

If you have software installed on the same zone as the oracle instance and you wish them to use oracle you must make sure to have ORACLE_HOME and ORACLE_SID set in there environment. In addition, you must add them to the oinstall group.

usermod -G oinstall username

Installing Apache

Building and installing Apache in a zone is very straight forward. Just remember in a sparse-root zone you cannot WRITE to /usr/local/bin so you must install configure Apache to install elsewhere, like /opt/apache2

After downloading and unpacking Apache, enter the following commands

./configure --prefix=/opt/apache2
make
make install

See the following for more information:
http://httpd.apache.org/docs/2.0/install.html

Conclusion

Hope you fond some of this helpful!

Workgroup Authentication

mike.clarke — Mon, 21 Jul 2008 14:04:00 GMT

Workgroup Authentication

Before a Windows XP Professional workstation joins a domain, the local security policy is set to Guest only - local users authenticate as Guest. This will prevent you from accessing UNC shares across the network.

This is the default setting is for increased security.

Switching Back to Classic

In order to allow the local machine to authenticate local users, you need switch the Sharing and security model for local users back to Classic - local users authenticate themselves

To do this, open up the Local Security Policy
Start > Control Panel > Administrative Tools > Local Security Policy

Set the following:

Done!

IIS Trace Logging in Windows Server 2003 SP1

mike.clarke — Mon, 23 Jun 2008 15:44:00 GMT

IIS Trace Logging in Windows Server 2003 SP1

This post will briefly describe how to enable IIS trace logging in IIS for Windows Server 2003 SP1. Note, that trace logging for IIS is NOT available in Windows XP

Here is the official Microsoft documentation which provides more details than this post will cover
http://www.microsoft.com/technet/prodtechnol/WindowsServer2003/Library/IIS/5f5bd256-7d1f-4239-9a7f-8eea4072fcb3.mspx?mfr=true

Logman

The first step in performing trace logging is to enable logging of the providers you wish to monitor. The tool to perform logging with is logman. See
http://technet2.microsoft.com/windowsserver/en/library/25d92f21-ffad-45c7-824e-b8c291559ebd1033.mspx?mfr=true

Basically, you invoke logman with a session name and a list of providers. The logging will then take place based on when you schedule logman to run. Logman will then produce a .etl (Event Trace Log). The final step is to parse this log into something you can read.

Providers

Here is a list of all the providers in Windows Server 2003 SP1 and greater:

C:\>logman query providers

Provider                                 GUID
-------------------------------------------------------------------------------
DNS Trace                                {1540ff4c-3fd7-4bba-9938-1d1bf31573a7}
IIS: WWW Global                          {d55d3bc9-cba9-44df-827e-132d3a4596c2}
ACPI Driver Trace Provider               {dab01d4d-2d48-477d-b1c3-daad0ce6f06b}
Active Directory: Kerberos               {bba3add2-c229-4cdb-ae2b-57eb6966b0c4}
IIS: SSL Filter                          {1fbecc45-c060-4e7c-8a0e-0dbd6116181b}
IIS: Request Monitor                     {3b7b0b4b-4b01-44b4-a95e-3c755719aebf}
IIS: WWW Server                          {3a2a4e84-4c21-4981-ae10-3fda0d9b0f83}
IIS: Active Server Pages (ASP)           {06b94d9a-b15e-456e-a4ef-37c984a2cb4b}
Local Security Authority (LSA)           {cc85922f-db41-11d2-9244-006008269001}
IIS: IISADMIN Global                     {DC1271C2-A0AF-400f-850C-4E42FE16BE1C}
Processor Trace Information              {08213901-B301-4a4c-B1DD-177238110F9F}
Windows Kernel Trace                     {9e814aad-3204-11d2-9a82-006008a86939}
ASP.NET Events                           {AFF081FE-0247-4275-9C4E-021F3DC1DA35}
NTLM Security Protocol                   {C92CF544-91B3-4dc0-8E11-C580339A0BF8}
IIS: WWW Isapi Extension                 {a1c2040e-8840-4c31-ba11-9871031a19ea}
Active Directory: SAM                    {8e598056-8993-11d2-819e-0000f875a064}
HTTP Service Trace                       {dd5ef90a-6398-47a4-ad34-4dcecdef795f}
Active Directory: NetLogon               {f33959b4-dbec-11d2-895b-00c04f79ab69}
Spooler Trace Control                    {94a984ef-f525-4bf1-be3c-ef374056a592}

The command completed successfully.

C:\>

We are interested in the following providers:

IIS: WWW Server
IIS: SSL Filter
IIS: ISAPI Extension
IIS: ASP
IIS: IISAdmin

See:
http://technet2.microsoft.com/windowsserver/en/library/25d92f21-ffad-45c7-824e-b8c291559ebd1033.mspx?mfr=true

As you can see, these providers are not available in Windows XP (any SP1)

C:\>logman query providers

Provider                                 GUID
-------------------------------------------------------------------------------
HTTP Service Trace                       {dd5ef90a-6398-47a4-ad34-4dcecdef795f}
WSAT Trace Provider                      {7f3fe630-462b-47c5-ab07-67ca84934abd}
ASP.NET Events                           {AFF081FE-0247-4275-9C4E-021F3DC1DA35}
Windows Kernel Trace                     {9e814aad-3204-11d2-9a82-006008a86939}
Processor Trace Information              {08213901-B301-4a4c-B1DD-177238110F9F}
ACPI Driver Trace Provider               {dab01d4d-2d48-477d-b1c3-daad0ce6f06b}
.NET Common Language Runtime             {e13c0d23-ccbc-4e12-931b-d9cc2eee27e4}

The command completed successfully.

C:\>

Getting it Done

The easiest way to get logging done is through batch scripting. For now, you will need three files:

start_log.bat
stop_log.bat
providers.txt

providers.txt

providers.txt will contain a list of the providers that you wish to log. Providers are defined in the following way:

"provider" flags [(0) is none] verbosity level

In this example, I will use the following for providers.txt:

"IIS: WWW Server" 0xFFFFFFFE 5
"IIS: Active Server Pages (ASP)" 0 5
"IIS: WWW Isapi Extension" 0 5

start_log.bat

We will use start_log.bat to start logman using the providers in providers.txt. In addition, must name this session so can stop it later. I've called my session demo

Here is what start_log.bat looks like:

logman start demo -pf providers.txt -ets

So now tracing is running. You see a new file called demo.etl created. That is the log. If your troubleshooting something, now would be a good time to try and reproduce the problem.

stop_log.bat

stop_log.bat stops the logging session you got going.

Here is what my stop_bat.log looks like. Recall, my session name is demo

logman stop demo -pf providers.txt -ets

demo.etl

So logman creates a .etl with the name of the session. In this example, the file is called demo.etl. This file that is the trace log. You can't read in. In order to read it you need to parse it into something you can read. I'll cover two methods for parsing the .etl file into a human readable format.

Logparser

The easiest way to few the trace logs is to use logparser, a tool included in IIS Diagnostics Toolkit. This can be downloaded here:
http://www.microsoft.com/downloads/details.aspx?familyid=9BFA49BC-376B-4A54-95AA-73C9156706E7&displaylang=en

The diagnostic toolkit includes a VB script that will parse the .etl file into text. Use the following command create demo.txt that will contain the complete trace log.

"%systemroot%\System32\cscript.exe" "C:\Program Files\IIS Resources\Log Parser 2.2\Samples\Scripts\DumpTraceReqs.js" demo.etl > demo.txt

Tracerpt

The alternative method is to use Tracerpt. See
http://technet2.microsoft.com/windowsserver2008/en/library/cb9eaf86-0ef6-4197-b6c8-9cca8a1d723c1033.mspx?mfr=true
http://technet.microsoft.com/en-us/library/bb490959(TechNet.10).aspx
http://www.microsoft.com/technet/prodtechnol/WindowsServer2003/Library/IIS/594174a1-f8d4-4338-a996-9c171e7bc3a2.mspx?mfr=true

The default outputs from Tracerpt do not seem as useful. You can pile on the switches to make it more useful, see Additional Resources. You can invoke the parsing with the following command:

tracerpt demo.etl

Conclusion

Hope you found this information useful!

Additional Resources

Trouble Joining the Domain

mike.clarke — Wed, 11 Jun 2008 16:05:00 GMT

Trouble Joining the Domain

Ever had a PC that just refused to join the domain? I know I have!

I've found the following two MS articles extremely help:

This article is about how to turn back on SMB signing, it applies (usually) when you are trying to add a computer that belonged to another domain:
http://support.microsoft.com/kb/908372

This article is how RPC calls changed from Windows 2003 to SP1 and how some firewalls block new RPC. It goes on to explain a registry hack to revert back to straight 2003:
http://support.microsoft.com/kb/899148

Getting Started with Solaris 10

mike.clarke — Wed, 21 May 2008 04:35:00 GMT

Getting Started with Solaris 10

The following covers varies topics related to getting started with Solaris 10

Configuring an IPv4 Network Interface in Solaris 10

When installing Solaris 10, you can configure a NIC a during the installation or you can configure it post install by changing these files:

/lib/method/net-physical
/etc/hostname.interface
/etc/inet/hosts, /etc/inet/ipnodes
/etc/inet/netmasks
/etc/defaultdomain
/etc/defaultrouter
/etc/nodename
/etc/resolv.conf
/etc/nsswitch.conf

/lib/svc/method/net-physical

In short, the script plumbs each /etc/hostname.interface. It effectively installs the interface and enables the kernal to communicate with the named network interface. No changes need to be made to this file.

/etc/hostname.interface

This file defines the network interface on the local host. This file contains only one entry: the hostname or IP address associated with the network interface.

/etc/inet/hosts, /etc/inet/ipnodes

Database that maps IP addresses to DNS names. Each line in the /etc/inet/hosts file uses the following syntax:

address hostname nickname #comment

/etc/inet/ipnodes is the same as hosts except ipnodes can support IPv6 addresses.

/etc/inet/netmasks

Contains the subnet mask for the network. The netmasks file is formatted like this

ip-range subnet

/etc/defaultdomain

This file should contain one entry, the full qualified domain name of the domain to which the local host's network belongs.

/etc/defaultrouter

This file should contain an IP address/hostname for each router directly connected to the network.

/etc/resolv.conf

Used to configure the DNS settings. The file supports the following entries:

domainname - the fully qualified domain name
nameserver - IP address of the first name server
nameserver - IP address of the second name server
nameserver - IP address of the third name server
search - prefixess to append to hostnames when trying to resolve name

Example:

domainname lab.local
nameserver 192.168.0.2
search lab.local

Useful Commands

netstat -rn

Displays the routing table.

netstat -f

Flushes the routing table.

Installing VMware Tools

Correcting Network - hostname.pcn0

After installing VMware tools on a i386 VM running Solaris 10, you may have to rename hostname.pcn0 to hostname.vmxnet0.

Correcting Screen Resolution - /etc/X11/xorg.conf

Edit /etc/X11/xorg.conf and remove all modes except the desired ones.

root Remote Access

By default the root can only log into the server console. Doing the following will grant the root remote access.

telnet - /etc/default/login

Comment out the following line to allow the root to login using telnet
CONSOLE=/dev/console

ssh - /etc/ssh/sshd_config

Set PermitRootLogin to yes to allow the root to login through ssh
PermitRootLogin yes

ftp - /etc/ftpd/ftpusers

Delete the following line from excluded ftp users
root

Remote X Sessions

Ensure the following default settings in /etc/ssh/sshd_config:
X11Forwarding yes
X11DisplayOffset 10
X11UseLocalhost yes
Download and install Xming from:
http://sourceforge.net/projects/xming
Start XLaunch and choose the following:
Choose Start a Program:
Select/Type gnome-session and hostname and credentials for connection:
Share clipboard, Next:
Finish:
Done!

Hope you found some of this useful!

References

Recovering A MSCS Cluster From A Failed Quorum Drive

mike.clarke — Fri, 02 May 2008 21:09:00 GMT

Recovering A MSCS Cluster From A Failed Quorum Drive

In my experience, Quorum drive failures are by the scariest as finding resources on this very specific problem is difficult. The following is a rough guide to recovering a MSCS cluster from a failed Quorum drive.

Starting the Cluster Service

With the Quorum disk gone, the service wouldn't start. If the service wouldn't start, you can't configure/fix the cluster.

The trick is to start the Cluster Service with the following parameter:
/fixquorum

Create a Cluster Disk Resource

Now that you have the cluster service running, go into Cluster Administrator and create a new disk resource in the Cluster Group group. This new disk is going to replace the failed Quorum disk so it must meet all the requirements of the original quorum disk.

NOTEYou either need to use . or the hostname of the node to get into Cluster Administrator. You cannot open Cluster Administrator using the Cluster Name. At this point, the cluster is still down.

Take this disk offline

Swap the failed disk for the new Disk

Traditionally if the Quorum disk failed you had to manually hack the registry to replace the drive. See:
http://support.microsoft.com/kb/243195

Fortunately, there is a tool available from Microsoft to perform all the registry changes to switch the Quorum disk. See
http://technet2.microsoft.com/WindowsServer/en/Library/b410b421-78a5-4b3f-9247-e4f248f878fc1033.mspx?mfr=true

The tool is called ClusterRecovery.exe and can be downloaded as part of the Windows Server 2003 Resource Kit located here:
http://www.microsoft.com/downloads/details.aspx?FamilyID=9D467A69-57FF-4AE7-96EE-B18C4790CFFD&displaylang=en

Alternatively, you can downloaded ClusterRecovery.msi which installs ClusterRecovery.exe by itself here:
http://www.microsoft.com/downloads/details.aspx?FamilyID=2BE7EBF0-A408-4232-9353-64AAFD65306D&displaylang=en

After you've installed ClusterRecovery.exe, run it by going Start > Run > ClusterRecovery.exe
You still need to use either . or the hostname of the node to connect to the cluster service. Complete the forms to swap the Quorum disk with the new Disk Resource.

Complete the Swap

Restart the cluster service without the /fixquorum parameter. Start all the other nodes. Delete the failed quorum disk.

Done!

Quorum Disk Resources

Microsoft Server 2008 Certification Upgrade Guide

mike.clarke — Fri, 18 Apr 2008 19:28:00 GMT

Microsoft Server 2008 Certification Upgrade Guide

Before reading this guide, it is helpful to have an understanding of the new certifications available for Microsoft Windows Server 2008. I recommend reading the following prior to reading this guide:
http://mystyleit.com/blogs/mystyleit/archive/2008/02/06/microsoft-server-2008-certification-guide.aspx

This guide will not cover all the upgrade paths, only the MCSA and MCSE 2008 equivalents. All upgrade paths can be found here:
http://www.microsoft.com/learning/mcp/mcts/windowsserver/2008/transition/default.mspx
http://download.microsoft.com/download/f/9/f/f9fa4e7c-ba7b-46f2-8002-3a60121a6d81/WinServer2008_transition_exam_roadmaps.pdf

Upgrading a Microsoft Windows 2003 Server MCSA

Recall, the 2008 MCSE equivalent is MCITP: Server Administrator. If you have the MCSA certification for Windows Server 2003 you can stream line the following MCTS requirements for MCITP: Server Administrator:

MCTS Prerequisite Exams

Number	Name
70-640	TS: Windows Server 2008 Active Directory, Configuring
70-642	TS: Windows Server 2008 Network Infrastructure, Configuring

MCTS Prerequisite Upgrade Exam

Microsoft has created a special upgrade exam that combines 70-640 and 70-642 into a single exam.

Number	Name
70-648	TS: Upgrading Your MCSA on Windows Server 2003 to Windows Server 2008, Technology Specialist

Professional Series Exam

You still must complete the Professional series exam to earn the credential.

Number	Name
70-646	PRO: Windows Server 2008, Server Administrator

MCSA to MCITP: Server Administrator Upgrade Summary

Upgrading a Microsoft Windows 2003 Server MCSE

Recall, the 2008 MCSA equivalent is Server Enterprise Administrator. If you have the MCSE certification for Windows Server 2003 you can stream line the following MCTS requirements for MCITP: Server Enterprise Administrator:

MCTS Prerequisite Exams

Number	Name
70-640	TS: Windows Server 2008 Active Directory, Configuring
70-642	TS: Windows Server 2008 Network Infrastructure, Configuring
70-643	TS: Windows Server 2008 Applications Infrastructure, Configuring

MCTS Prerequisite Upgrade Exam

Microsoft has created a special upgrade exam that combines 70-640, 70-642 and 70-643 into a single exam.

Number	Name
70-649	TS: Upgrading Your MCSE on Windows Server 2003 to Windows Server 2008, Technology Specialist

Additional MCTS Requirements

You still must complete one of the following additional MCTS exams.

Number	Name
70-620 OR 70-624	TS: Windows Vista, Configuring OR TS: Deploying and Maintaining Windows Vista Client and 2007 Microsoft Office System Desktops

Professional Series Exam

You still must complete the Professional series exam to earn the credential.

Number	Name
70-647	PRO: Windows Server 2008, Enterprise Administrator

MCSE to MCITP: MCITP: Enterprise Administrator Upgrade Summary

This concludes the guide. Good luck on those exams!

Adding Social Bookmarking to Community Server 2007

mike.clarke — Thu, 07 Feb 2008 18:05:00 GMT

Adding Social Bookmarking to Community Server 2007

Community Server is the platform that powers rich blogging, discussions, and sharing web communities. It is widely used by IT Professionals for personal blogs. mystyleit.com for example is powered by Community Server.

This guide will summarize how to add Social Bookmarking to Community Server using the CS module developed by Christopher Even of sharepointsearch.com.

Here is the link to Christopher's original post:
http://sharepointsearch.com/cs/blogs/site_admin/archive/2007/09/15/adding-refer-digg-technorati-links-to-community-server-blogs.aspx

Installing the Module

Christopher Even has published module for download. The module can be found here:
http://sharepointsearch.com/cs/blogs/site_admin/attachment/2349.ashx

Download the file and extract the DLL from the ZIP file and copy it your Community Server's \bin folder.

Next, you will need to make an entry for the module in your CommunityServer.config file. To do this, add the following line to CommuntyServer.config at the end of the CSModules element. Around line 1078.

<add name = "SPWReferItModule" type = "SPWorks.CS.Freeware.SPWReferIt,SPWorks.CS.Freeware" ReferLinks="template" />

Take notice of template variable, this will come up later.

Configuring the Module

If you can follow the code is Chris's module, you'll see that the module pulls a template string from the CommunityServer.config and then does the following:

Replaces all pre-defined tags in the template string with page specific values
Writes out the string at the end of your blog posts

Originally Chris intended for users of his module to create their custom social bookmarking links for their blogs. But there are several ways to create the links. I'll cover the three options I see.

Intended Use

Chris's intended use was for users of his module to configure their own links using his pre-defined tags.

For example, here are links for del.icio.us and digg:


<a href="http://del.icio.us/post?url=[URL_ENCODED]&tags=[TAGS_ENCODED]&title=[TITLE_ENCODED]" mce_href="http://del.icio.us/post?url=[URL_ENCODED]&tags=[TAGS_ENCODED]&title=[TITLE_ENCODED]">Del.icio.us< /a>


<a href="http://digg.com/submit?phase=2&url=[URL_ENCODED]&title=[TITLE_ENCODED]&tags=[TAGS_ENCODED]" mce_href="http://digg.com/submit?phase=2&url=[URL_ENCODED]&title=[TITLE_ENCODED]&tags=[TAGS_ENCODED]">Digg It< /a>

Notice tags in the strings. The module will populate these with the correct values. You can fancy it up a bit, notice on Chris's page he has added a | between his text links.

Once you've figured out how to link to your favourite sites and created your string, you need to URL Encode it before you can plunk it in as the template variable.

You can use the following website to URL encode your string:
http://www.albionresearch.com/misc/urlencode.php

Once your done, you'll have links similar to those on Chris's blog.

Intended Use With a Twist

In another follow up post it was pointed out by David Leibowitz of jetlounge.net that these are just links and there is no reason these can't be icon links.

Here is the link to David's post:
http://jetlounge.net/blogs/teched/archive/2007/12/30/add-referring-icons-to-your-community-server-posts.aspx

David provide the following as an example in his post:


<a href="http://del.icio.us/post?url=[URL_ENCODED]&tags=[TAGS_ENCODED]&title=[TITLE_ENCODED]" mce_href="http://del.icio.us/post?url=[URL_ENCODED]&tags=[TAGS_ENCODED]&title=[TITLE_ENCODED]" target="_blank"><img class=link_icon alt="del.ico.us" src="/images/delicious.png" alt="del.ico.us" /></a>

As you can see, has simply changed link from a text link to a image link. He stores the images of his favourite social bookmarks in /images.

Besides the type of link, the procedure is exactly the same. Construct the string and URL encode with http://www.albionresearch.com/misc/urlencode.php then plunk it in as the template variable.

Once your done, you'll have links similar to those on David's blog.

Another Use

Creating the links is time consuming. Even with the pictures it still doesn't' look great.

Because Chris's module just writes out whatever is in the template variable, you can basically plunk in whatever you want in there. This bring us my use of Chris' module.

I choose to sign up for an addthis.com widget. After you sign up and get your sourcecode from them, all you need do is URL encode using http://www.albionresearch.com/misc/urlencode.php and use the result as the template variable. The advantage of an addthis.com widget is that addthis.com tracks if people are actually clicking the buttons. Something the previous don't offer. The disadvantage is sometimes it can take a while to load.

I think the result looks a lot better. I believe you could use the module to deploy any social widget you wanted, like the digg widget.

Hope you enjoyed my guide.