Exception Handling and Logging – Using Microsoft.Practices.EntrepriceLibrary

System.Security.SecurityException: The source was not found, but some or all event logs could not be searched. Inaccessible logs: Security."

I received this error when I was trying to log application exception details to the event log using below method. With bit of googling I understood that my app is trying to write to the Event Log using a value for "Source" that has not been registered. As the error says it is a security issue, but I was wondering what it is, because my code tries to create the source if it is not available and I am running this application with administrative privileges (my domain account is added to local administrators group).

(My dev environment is VS 2008 running on Win Server 2008)

/// <summary>
/// Logs the message to the EventLog.
/// </summary>
private static void LogEvent(string message, 
            System.Diagnostics.EventLogEntryType eventLogEntryType)
{
   // Log available in EventLog.
   const string LOG_APPLICATION = "Application";
   const string LOG_SOURCE = "Project1.Log.Test";

   if (!System.Diagnostics.EventLog.SourceExists(LOG_SOURCE))
   {
     System.Diagnostics.EventLog.CreateEventSource(LOG_SOURCE, LOG_APPLICATION);
   }

   System.Diagnostics.EventLog.WriteEntry(LOG_SOURCE,message,eventLogEntryType);
}

So I tried executing "EventSourceInstaller.exe" again from my BIN directory and I noticed the same error.

To overcome this issue I had to run EventSourceInstaller with administrator account of the server and now all working fine. I guess this has something to do with Win Server 2008 but not with Win Server 2003. Here are the basic steps I had to follow for the purpose of logging with Enterprise Library.

• Add references to the enterprise library dlls
  
• Add necessary configuration entries to the Web.Config
  
• Run 'EventSourceInstaller.exe' from application bin directory (With administrator account) – Make sure source value is same in Web.Config and EventSourceInstaller.exe.config are same
  
• Use System.Diagnostics.EventLog.WriteEntry(…..) for logging
  

Prism - Mozilla Labs

Prism is a new project by Mozilla labs which is capable of separating web applications from the browser and running Web Apps as separate desktop applications. In other words a Web App can be run as a desktop application and not run from within a browser. Also in some cases Prism can be used even to run Web applications offline. With Prism applications run separately from the browser, so they stay up even if the browser doesn't.

Prism is a web 2.0 application, come as a standalone application as well as a FireFox extension. It is not a new platform, but the web platform integrated with desktop experience. Therefore Web developers don't have to target Prism separately, as any application that can run in a modern standard web browser can run in it and it supports technologies like HTML, JavaScript, CSS.

Main Features of Prism

• Access web apps from system taskbar
• Tray icon and dock menus
• Run applications as startup apps
• Associate applications with browser links
• Minimize to tray
• Popup alerts

Both flavors of Prism can be downloaded from Mozilla Labs here.

Determine SQL Server Object Dependencies with sp_Depends and sp_ MSdependencies

The sp_Depends system stored procedure displays information about object dependencies for a table, view, stored procedure, user defined function, or trigger using the information stored in a system table named sysdepends.

Syntax:
EXEC sp_depends @objname = N'dbo.table1';

This will return both the name and type of the objects dependent on 'table1'. However sp_depends doesn't list tables, but does list check constraints.

The sp_MSdependencies can also be used to find dependent objects. In here, depending on the specified parameters the results will be varied. For example;

Syntax: EXEC sp_MSdependencies N'dbo.table1', null, 1315327

Syntax:EXEC sp_MSdependencies N'dbo.table1', null, 1053183

Syntax: EXEC sp_MSdependencies N'dbo.table2', null, 1315327

Syntax:EXEC sp_MSdependencies N'dbo.table2', null, 1053183

Syntax: EXEC sp_MSdependencies N'dbo.table1'

Syntax: EXEC sp_MSdependencies N'dbo.table2'

Syntax: EXEC sp_MSdependencies N'dbo.table3'

Difference between Clustered and Non-Clustered Indexes

Clustered Index: index entries are actually data records and the database table is physically sorted on the basis of that index. Therefore when the database searches for an entry, it will end up with the actual data a lot faster. There can be only one Clustered index for a table. So this is usually made on the primary key.

Clustered indexes are good for:

• Columns that contain a large number of distinct values
• Queries that return a range of values using operators such as BETWEEN, >, <, >=, <=
• Columns that are accessed sequentially
• Queries that return large result sets

Non- Clustered Index: Index entries are not actually data records and index is UNSORTED. In here logical order of the index does not match with the physical stored order of the rows on the disk. That means the index data is stored outside the table and contains a sorted reference to the table. There can be many Clustered indexes for a table.

Ex: Clustered index is like a phone directory, where actual records are sorted by name. Non- clustered index is like an 'Index' in the end of a book, where actual records are not sorted. But you have sorted references to the book content.

Note: Clustered Index is an SQL Server term and is not like an Oracle cluster, but more like an Oracle Index Organized Table (IOT).

Indexing the table makes the data retrieval faster but it has a disadvantage that the insert and update statements start to take longer as all the indexes are updated and this effect is different in the two approaches. Because in clustered index you'll have to split the pile in the middle and insert/update records. In non-clustered it is only updating sorted references.

NTFS File System for USB Flash Drives

Recently I came across a situation where I wanted to copy a virtual hard disk (which is about 12GB in size) to my USB drive. Though there was enough free space on the disk I wasn't able to copy the file. Reason for this was maximum file size allowed in FAT file system is 4096MB.

So I thought of formatting my flash drive with NTFS file system and just searched for pros and cons.

Pros:

• Avoid file size restriction
• NTFS allow file compression
• Manage permissions for individual files and folders
• Use windows XP built in file encryption

Cons:

• Additional data writes – NTFS logs disk transactions separately on the disk adding considerable amount of extra disk activity, which leads to wearing out your USB drive faster
• Windows 98/ ME, and most Linux systems cannot read NTFS partitions
• Must go through removal dialog, or else you have a good chance of losing data as you have to switch from the 'Optimize for quick removal' mode

Therefore it is up to you to choose between formatting flash drive with NTFS or not. As I read, if you have done your research and decided to format your flash drive with NTFS, here is how to do it.

• Right Click My Computer-> Select Manage-> Open Device Manager
• Select your USB Drive-> Right Click-> Select Properties-> Select Policies Tab -> Select Optimize for performance option-> Click OK

• Open My Computer-> Right Click on your USB Drive-> Select Format-> Select File System as NTFS-> Start

A good article on "Format a USB drive with NTFS file system" can be found here.