Hello. In this video, I will show you how to quickly isolate deadlock issues using foglight for databases. A deadlock is a situation when one request is waiting for another resource to complete the process. Once SQL server detects any deadlock, the foglight [? agent ?] captures the data. This information can be seen from deadlock's UI. This can be accessed from the database's home screen and then Select Activity, Deadlocks.
There are four tabs in the Deadlock UI page. The first tab displays all deadlocks for the selected time period. The Databases tab displays database-specific deadlocks. Similarly, Objects and Application tabs filter deadlocks specific to an object and application. The Deadlock table can be used to view SQL text, database, application name, lost time, time of last transaction, object in question, and object in a waiting state. Each line can be selected to view deadlock details or select download [? SSMS ?] file to view deadlock XML code.
The deadlock code can be opened in any XML [? data. ?] By looking at the XML code, you can determine which statement is terminated as deadlock victim, all statements involved in this deadlock, and if it's in a procedure, the name of the procedures involved, and so on. The downloaded file can be opened in Microsoft SQL server Management Studio.
The left oval on the deadlock graph with blue cross represents the transaction chosen as the deadlock victim. The right oval represents the successful transaction. The two middle rectangular boxes are called resource nodes, representing a database object such as table, row, or an index. The resource nodes represent indexes which each process was trying to get an exclusive lock on. The historical deadlocks and blocking can also be viewed from SQL PI from a multi-dimensional aspect.
For more information on blocking and deadlocks, please go to www.quest.com/foglight-for-databases/