I have a bunch of SQL Agent jobs that execute T-SQL against databases. A bunch of these databases are mirrored. Of course if the database is the principal then these jobs will work without a problem. But what happens if you failover? Now these jobs will…
Less Than Dot is a community of passionate IT professionals and enthusiasts dedicated to sharing technical knowledge, experience, and assistance. Inside you will find reference materials, interesting technical discussions, and expert tips and commentary.
I have been using mirroring for about three years now and I must say it is one of the best features that have been added to SQL Server. We used to do plain old replication and log shipping in the past but almost all of those have been replaced by mirror…
Since mirroring has come to the SQL Server feature set, High Availability (HA) has become a much easier, inexpensive and available option. Prior to mirroring, third party software and hardware were the main options to achieve HA. This was due to replication never being accepted as a high availability option. With any solution like SQL Server, growth over time builds a better and more stable solution. This has been the fact with many features of SQL Server, including replication. Replication had some issues in previous versions of SQL Server. Maintaining the uptime and stability was not a trivial task. With replication, there is a more in-depth knowledge requirement to maintain it, while mirroring does not require as much in-depth knowledge to provide a solid solution. Do think that mirroring does not include a strict discipline of skills in order to make it a successful HA solution. Mirroring is simplistic from the outside, both in setup and maintainability. However, troubleshooting mirroring is a critical aspect to making it work for HA. In order to be successful, knowledge of the internals of mirroring is as important as replication.
Realistically, 100% is unachievable given the nature of computing. There are needs for a SQL Server and Windows Server to be rebooted at least once a year. This is to allow for updates on both SQL Server and Windows to be maintained. So the ranking method we use for measuring high availability is the “nines” scale. The five nines is a goal that most database administrators and teams set for their standards. The five nines level is a height of availability that is truly an achievement and one to be proud of.
The recess bell just rang for SQL University HA / DR classrooms. While all of the SQL kiddies are running around the playground and playing with the things they have learned over this semester, the chalkboard is going to get a workout so when they get back, they can take the notes they slacked on earlier.
Welcome to day three of HA and DR week of SQL University. Today we are going to look at cheap DR. Yes, setting up DR can be inexpensive. The best part of this strategy is it comes along with most of the editions of SQL Server. The method is Log Shipping. Log shipping (LS) has a bad name in the Disaster / Recovery (DR) world. There are concerns with the ability to fail back to primary sites in the case of disasters, and LS is often thought of as a maintenance intense setup along with file mess. Today’s class will go over some methods to handle these and other concerns, along with the simplicity of configuring and monitoring LS in SQL Server 2008 (R2).
Welcome to our second class of HA and DR week of SQL University. Today we are going to focus on the concept, “Backups are for sissies!” OK, we’re really going to look at backup and restore for Disaster / Recovery (DR) and how being a sissy and always backing up our databases and testing out restores is a proven strategy for DR. When all else fails and the walls are falling down on the database servers, backups will be your life preserver. Backups are the foundation for Disaster and Recovery (DR). Backups can also save you when high Availability (HA) completely fails you. Let’s get started!
Time to get Dirty! Today we are going to get down into actually configuring a basic mirror using Developer Edition. Developer Edition is a great tool that is extremely inexpensive.
Hardware can be a single point of high performance and a single point of failure. In a mirroring situation, there is not a listing we can put on paper as to the best hardware and configuration we can make. Each database server is configured as it requires for IO operations, memory usage etc… RAID 1+0 would give you write performance on the mirror but at the cost of what if you database server that is active is on RAID 5?
Researching and obtaining the knowledge and test cases for configuring mirroring is a large part in putting SQL Server mirroring to work for your High Availability (HA) solution. Before you start jumping into configuring mirroring, several questions should be researched and answered. This ensures that the HA solution and mirroring will work for your environment and your business. One of the highest achievements in mirroring is to ensure it will always be invisible to your user community. After all, as DBAs, our position is one that truly is successful when the phones are not ringing and users forget who we are. Our positions have little to no recognition from their view. It’s a hard truth that we most of the time are far outside the spotlight, but one that is key to the role.