My torn data pages - June, 2007

Yesterday I enjoyed the webcast by Hariharan Sethuraman and Chris Haslam, both from Microsoft. They talked about ICE, which stands for Information Security Consolidated Event Management System. ICE collects the in- and outbound e-mail traffic, login events and web browsing (web proxy and firewall logs) and stores it for 60 days, to provide an audit-trail in case of security events. The webcast was about how ICE 3.0 was designed and built on top of the infrastructure below.

Imagine the numbers; 40 TB designed -- 27 TB allocated (FibreChannel SAN-)storage, designed to load 60 GB/hour into staging tables -- currently receiving max 1.2 TB daily with 600 GB as daily average, table partitioning, 4 (multi-core?) x64 processors & 32 GB RAM for the Database Engine + 4 (multi-core?) x64 processors & 8 GB RAM for the Integration Services. Accessing all that data via Ad-hoc queries and Report Server reports.

Already ICE version 4.0 is envisioned, however again only for internal usage, ICE is not (yet?) planned as a commercial product. So with ICE not being planned as a product, I just had to ask "Sounds like a great reference project for SQL Server 2005 and is a very useful application, are you planning a white-paper?"... Turned out the webcast is a precursor to the white-paper!!!

I'm looking forward to update this post with the link to that white-paper. For now I can only point you to the on-demand webcast: How Microsoft IT Uses SQL Server 2005 to Power a Global Forensic Data Security Tool (Level 300) (~60 minutes)

/* UPDATE 2007-11-22 */ Link to the afore mentioned white-paper.

Earlier I posted about pingback spam and a little tool I created to remove the spamvertised links, today I upgraded the tool from version 1.0.0.0 to version 1.0.0.1. So what's changed;

• The updated version doesn't leave empty Tracking elements after removing the spamvertied entries. From the dasBlog Eventlog I suspected that too many empty Tracking elements were causing errors in the trackbackList macro.
• An XML file is only written back if something changed in the file.

Source and executable files have been changed.

--- So how does the tool work? ---

First create a config file

Look in your dayfeedback-files for the spamvertised PermaLinks (or you may already know about them if you have Notification Settings enabled for pingback and trackback). It's the PermaLink pattern that matters:

<Tracking>
   <TargetTitle>Today's bad start</TargetTitle>
   <TargetEntryId>70e036a5-ce3f-4fa0-9463-816983848a48</TargetEntryId>
   <RefererTitle>http://9uams-le-informazioni.info/52600366/adempimenti-trasformazione-ditta-individuale.html</RefererTitle>
   <RefererExcerpt />
   <PermaLink>http://9uams-le-informazioni.info/52600366/adempimenti-trasformazione-ditta-individuale.html</PermaLink>
   <RefererBlogName>http://9uams-le-informazioni.info/52600366/adempimenti-trasformazione-ditta-individuale.html</RefererBlogName>
   <TrackingType>Pingback</TrackingType>
</Tracking>
<Tracking>
   <TargetTitle>Today's bad start</TargetTitle>
   <TargetEntryId>70e036a5-ce3f-4fa0-9463-816983848a48</TargetEntryId>
   <RefererTitle>http://9uane-free-porn.info/91478222/weight-and-height-for-celebrities.html</RefererTitle>
   <RefererExcerpt />
   <PermaLink>http://9uane-free-porn.info/91478222/weight-and-height-for-celebrities.html</PermaLink>
   <RefererBlogName>http://9uane-free-porn.info/91478222/weight-and-height-for-celebrities.html</RefererBlogName>
   <TrackingType>Pingback</TrackingType>
</Tracking>

The pattern for the second Tracking entry is already included in the default config file, so you would only have to add the pattern for the first Tracking entry. To do so, you only need to add the following XML-element in the Configs section:

<Config Pattern="*-le-informazioni.info/*" />

With the updated config file, you can now run the tool to clean all XML-files in your content folder.

Next is figuring out why Akismet isn't working for me.

A little while ago Howard Dierking (MSFT, Certification Planner for developer stuff) shared his thoughts on doing exam development different. I liked the idea which aims to both lower the cost of building an exam and at the same time increase the quantity and quality of the questions. I added some feedback, so exams can be kept up to date, thus improving the quality of certification.

Recently I had a lot of pingback spam coming into my blog and I can tell you, it's more annoying than in your mailbox (Shift + Del and it's gone again). With dasBlog, it takes a little more effort, either via the admin interface or by deleting the spamvertised tarckings from the dayentry xml files. I even disabled trackback and pingback, temporarily!

Also I searched for solutions to this and came up with a several other affected blogs and akismet.com. With Akismet being the most promising, I tried, but spam kept coming in and Akismet was timing out (currently running dasBlog 1.9.7067.0.).

Error:
The external spam blocking service failed for pingback from http://9qk-information.info/56369818/index.html. Original exception: System.Net.WebException: The operation has timed out at System.Net.HttpWebRequest.GetRequestStream() at Subtext.Akismet.HttpClient.PostRequest(Uri url, String userAgent, Int32 timeout, String formParameters) at Subtext.Akismet.AkismetClient.SubmitComment(IComment comment, Uri url) at Subtext.Akismet.AkismetClient.CheckCommentForSpam(IComment comment) at newtelligence.DasBlog.Web.Core.AkismetSpamBlockingService.IsSpam(IFeedback feedback) at newtelligence.DasBlog.Web.Services.PingbackAPI.ping(String sourceUri, String targetUri)
while processing /ssb/PermaLink,guid,e63313c2-6f4d-434f-8a62-09f82a4b534c.aspx.

Hmm, if someone has some suggestions, I'd be happy to read your comment.

-----

And that was about the spam coming in, I also had a nasty amount already polluting my blog. I could resign, disable *back permanently and clan manually once more... or fire up Visual Studio and write a piece of newbie/n00b code. Well the newbie pulled it off, his first useful C# program. At least, it removed the spammings from my blog.

So if you want to get rid of;

• http://9??-free-porn.info/*
• http://9??-information.info/*
• or whatever disliked pattern of tracking permalinks on your dasBlog.

And you know that you try the followig stuff at your own risk and made a backup of your blog's content folder.

• Compiled file (console application, run without parameters to see what you have to do to get it working. Then make the config file and if needed, edit it.). Requires .NET Framework 2.0.
• Source-code (Visual Studio 2005, C#)

Disclaimer: It worked for me (dasBlog 1.9.7067.0.), it may not work for you and you use it at your own risk. However, feel free to comment with question, suggestions or other stuff.

<UPDATE date="2007-06-18">Version 1.0.0.1 released, learn more here</UDPATE>

Past weeks haven't been boring, if you look at the amount of database related news from Microsoft. And plenty of stages to let the world know it's still moving... BI Conference, Mix'07. And now TechEd.

At BI Conference, the new version of SQL Server was officially announced. At Mix'07 besides Silverlight there also were Jasper and Astoria. To see the presentations go to sessions.visitmix.com and search on the session code DEV18 for Jasper and XD006 for Astoria.

The binding word for Jasper and Astoria and Katmai is Enity Data Model. You may find several usefull video's on Channel9;

• ADO.NET Entity Framework: What. How. Why.
• Pablo Castro: Astoria Data Services
• Meet the Project Jasper Team
• Dr. Peter Chen: Entity Relationship Model - Past, Present and Future

Yep, a lot of developer stuff, but it will show the importance of EDM for data driven applications and thus access to the database. Also, judging from this press release, one of SQL Server 2008 main delivery goals is accessibility of data.

With TechEd started, there is also more and more info about Katmai, including this recorded webcast. Like expected, a lot of Entity Framework (includes EDM) here. Other very noteworthy additions; spatial data types (demo with Virtual Earth), policy based management, data  and backup compression, resource governor and a FILESTREAM data type. FILESTREAM is about storing metadata in SQL Server and the file itself on the file system, while maintaining a relation between them (including for BACKUP/RESTORE, I imagine something like with Full-Text Indexes on SQL Server 2005). Sounds like WinFS, doesn't it?

Also the first CTP's are out, based on the Developer Edition. Download the x86 and/or x64 versions here, after having registered.