Analysis of MS SQL Server for those who see it for the first time – Part 1

Ive recently encountered a problem – SVN went down on ubuntu server. I develop for Windows and I do not have much experience with Linux. I googled the errorwithout success. The error turned out to be the most typical one (the server unexpectedly closed the connection) and does not indicate anything. Therefore, it was necessary to go deeper and analyze logs/settings/rights/etc.

Finally, I figured out the mistake and found everything I needed, but I spent a lot of time. After solving this problem, I thought about how to reduce the uselessly spent hours and decided to write an article that will help people quickly get the understanding of the unfamiliar software.


The Lost Update Problem in Concurrent Transactions

The lost update problem occurs when 2 concurrent transactions try to read and update the same data. Let’s understand this with the help of an example.

Suppose we have a table named “Product” that stores id, name, and ItemsinStock for a product.

It is used as part of an online system that displays the number of items in stock for a particular product and so needs to be updated each time a sale of that product is made.


Counting references to a record in a table via Foreign Keys

I have recently needed to solve the task for my own purpose: to calculate the number of external records linked by a foreign key for each record in a table (File). The task was solved for the specific structure of the File table, but if necessary, the solution can be reworked to a universal one.

I’ll clarify that the solution was developed for an unloaded database, without millions of records and an every minute update, so there was not much concern about the performance.


Synchronizing database structure between applications

Anyone who has ever developed applications that use a database has probably faced the problem of updating the database structure when the application is deployed and updated.

The most common approach is to create a set of SQL scripts to modify the database structure from version to version. Of course, there are paid tools, but they do not always solve the problem of full automation of the update.


Understanding Dirty Read Problem with SQL Server

One of the most common problems that occur while running concurrent transactions is the Dirty Read problem. A dirty read occurs when one transaction is permitted to read data that is being modified by another transaction which is running concurrently but which has not yet committed itself.

If the transaction that modifies the data commits itself, the dirty read problem doesn’t occur. However if the transaction that modifies the data is rolled back after the other transaction has read the data, the latter transaction has dirty data that doesn’t actually exist. (more…)