In this blog post, we will discuss an issue encountered with an Always On Availability Group where a database remained stuck in the Initializing / In Recovery state on the secondary replica after a failover. We will explore the steps taken to resolve the problem and provide insights into the underlying causes.
Issue
During a failover from the primary to the secondary replica in an Always On Availability Group, one particular database entered the Initializing / In Recovery state on the secondary replica. Despite restarting SQL services and attempting to suspend data movement, the database remained stuck in this state.
Resolution
Logging off users and restarting SQL services: Initially, all users were logged off from the secondary replica, and the SQL services on the replica were restarted. However, these steps did not resolve the issue.
Suspending data movement: The next approach involved suspending the data movement of the problematic database from the SQL Server Management Studio on the secondary replica. However, this operation resulted in a “lock request time out period exceeded” error.
Identifying the blocking process: To investigate further, a query was executed to identify any blocking processes using the following query: SELECT * FROM sysprocesses WHERE blocked > 0 It was discovered that the DB Startup thread for the specific database was blocking the SPID attempting to suspend data movement.
Checking recovery status: The SQL Server error log was examined to determine if the database’s recovery had completed. The log indicated that recovery for the database had finished in 43 seconds, with no user action required. Recovery completed for database JB_DB (database ID 7) in 43 second(s) (analysis 64 ms, redo 324 ms, undo 0 ms.) This is an informational message only. No user action is required.
Analyzing DB Startup thread wait type: Monitoring the DB Startup thread revealed that it was in a wait state with the HADR_RECOVERY_WAIT_FOR_UNDO type. This wait type occurs when the secondary database is waiting for the reverting and initializing phases to synchronize with the primary database after failover.
Monitoring undo progress: The “Log remaining for undo” performance monitor counter was used to track the amount of log in kilobytes remaining for the undo phase. Surprisingly, the counter showed 0, indicating that there was nothing to undo, and the recovery completed message in the SQL Server error log confirmed that the undo took 0 milliseconds.
Failing over to another synchronous secondary replica: To further troubleshoot the issue, a failover was performed from the primary replica to another synchronous secondary replica. This time, all databases returned to a synchronized state without any issues.
Summary
The issue of a database stuck in the Initializing / In Recovery state after an Always On Availability Group failover was resolved by failing over to another synchronous secondary replica. Although the root cause and exact timing of the issue remain unknown, this blog post aims to provide guidance and assistance to others who may encounter similar challenges within the community.
Thank You, Vivek Janakiraman
Disclaimer: The views expressed on this blog are mine alone and do not reflect the views of my company or anyone else. All postings on this blog are provided “AS IS” with no warranties, and confers no rights.
SQL Server Accelerated Database Recovery (ADR) is a feature introduced in SQL Server 2019 that enhances the database recovery process. ADR addresses the long-standing challenge in SQL Server of database recovery time increasing with the number of transactions and active connections to the database. ADR helps minimize downtime during database recovery and reduces the risk of data loss.
This article will provide an in-depth look at SQL Server Accelerated Database Recovery, including its background, challenges and opportunities, best practices, future outlook, and conclusion.
Traditionally, when SQL Server experiences a crash or an unexpected shutdown, the database goes through a recovery process that can take a long time to complete. The recovery process involves three phases: analysis, redo, and undo. During the analysis phase, SQL Server scans the transaction log to determine the point at which the database was last consistent. In the redo phase, SQL Server applies all the transactions that were committed after that point. Finally, in the undo phase, SQL Server rolls back any transactions that were not committed.
The time it takes to complete the recovery process depends on the size of the transaction log, the number of transactions, and the number of connections to the database. In some cases, the recovery process can take hours or even days to complete, causing significant downtime for the application and potentially resulting in data loss.
Explanation of database recovery Whenever a SQL Server instance restarts, the databases that were online before the restart must undergo a recovery process. During the recovery process, SQL Server ensures that all committed transactions are written to disk and any uncommitted transactions are rolled back. This ensures that the database is brought back to a consistent state.
Overview of SQL Server Accelerated Database Recovery SQL Server Accelerated Database Recovery is a feature introduced in SQL Server 2019 that improves database recovery times and availability. It accomplishes this by reducing the amount of log data that needs to be replayed during recovery and by allowing recovery from a checkpoint.
Benefits of SQL Server Accelerated Database Recovery The primary benefits of SQL Server Accelerated Database Recovery are faster recovery times and improved availability. This feature significantly reduces the downtime associated with database recovery and allows organizations to restore their systems more quickly in the event of a failure.
Understanding Traditional Database Recovery
Traditional database recovery involves restoring the database from a backup and replaying all the transactions in the log since the last backup.
How traditional database recovery works In traditional database recovery, SQL Server reads the transaction log to identify all transactions that were not committed at the time of the restart. SQL Server then rolls back all uncommitted transactions and applies all committed transactions to the database. This process can take a significant amount of time, depending on the size of the database and the number of transactions that were in progress at the time of the restart.
Limitations of traditional database recovery Traditional database recovery has several limitations. It can take a long time to complete, especially for large databases. In addition, the recovery process can cause a significant amount of I/O activity, which can impact the performance of the server. Finally, if a failure occurs during the recovery process, the entire recovery process must be restarted.
How SQL Server Accelerated Database Recovery Works
Log Sequence Number (LSN) filtering SQL Server Accelerated Database Recovery works by filtering out redundant transaction log records during the recovery process. This is accomplished through the use of Log Sequence Number (LSN) filtering, which is a feature that was introduced in SQL Server 2019.
When a database is in Accelerated Database Recovery mode, SQL Server maintains a version store, which is a collection of active and previous versions of data pages. Each version of a data page is identified by its LSN, which is a unique identifier assigned to each transaction log record.
During the recovery process, SQL Server filters out transaction log records that are already reflected in the version store. This means that only changes that occurred after the most recent checkpoint are replayed during recovery, which can significantly reduce the amount of time required for recovery.
Checkpoint process Another key aspect of SQL Server Accelerated Database Recovery is the checkpoint process. Checkpoints are a mechanism used by SQL Server to write dirty data pages (i.e., data pages that have been modified but not yet written to disk) to disk. This helps to reduce the amount of work required during recovery, as it ensures that there is less dirty data to be written to disk when the recovery process begins.
With Accelerated Database Recovery, the checkpoint process is enhanced to include a special type of checkpoint called a Accelerated Database Recovery checkpoint. These checkpoints are optimized for use with Accelerated Database Recovery, and they help to ensure that the version store is properly maintained and that redundant transaction log records are filtered out during recovery.
Recovery with Accelerated Database Recovery enabled When a database is in Accelerated Database Recovery mode, recovery is performed in a slightly different way than it is with traditional database recovery. Instead of replaying all transaction log records from the beginning of the log, SQL Server uses the version store to filter out redundant records and only replays the necessary changes.
This can result in significantly faster recovery times, particularly for large databases or databases with high transaction rates. In addition, because only necessary changes are replayed, there is minimal impact on workload during the recovery process.
Benefits of SQL Server Accelerated Database Recovery
Faster recovery times One of the primary benefits of SQL Server Accelerated Database Recovery is faster recovery times. By filtering out redundant transaction log records and replaying only necessary changes, SQL Server can significantly reduce the amount of time required to recover a database.
This is particularly beneficial for large databases or databases with high transaction rates, as traditional database recovery can take a significant amount of time in these scenarios. With Accelerated Database Recovery, recovery times can be reduced from hours or even days to minutes.
Improved availability Another benefit of SQL Server Accelerated Database Recovery is improved availability. Because recovery times are significantly reduced, databases can be back up and running more quickly after a failure.
This can help to minimize downtime and ensure that critical business processes are not impacted by database failures. In addition, because only necessary changes are replayed during recovery, there is minimal impact on workload during the recovery process, which further improves availability.
Minimal impact on workload With traditional database recovery, there is a significant impact on workload during the recovery process. This is because all transaction log records must be replayed from the beginning of the log, which can result in significant resource usage.
With SQL Server Accelerated Database Recovery, only necessary changes are replayed during recovery, which significantly reduces the impact on workload. This means that business processes can continue to operate normally during the recovery process, which is particularly important for mission-critical applications.
Implementation of SQL Server Accelerated Database Recovery
Implementing SQL Server Accelerated Database Recovery is a straightforward process, but there are a few requirements to keep in mind.
Compatibility requirements First, your SQL Server instance must be running on at least SQL Server 2019 Enterprise Edition or Azure SQL Database. Additionally, your database must be running in the compatibility level 150 or higher.
Enabling Accelerated Database Recovery To enable Accelerated Database Recovery for a specific database, use the following T-SQL command:
ALTER DATABASE [DatabaseName] SET ACCELERATED_DATABASE_RECOVERY = ON;
Once enabled, Accelerated Database Recovery is applied to all operations performed against the database. This includes all transactional operations, such as inserts, updates, and deletes, as well as DDL operations, such as table creation or index rebuilding.
Monitoring Accelerated Database Recovery SQL Server provides several mechanisms for monitoring the performance of Accelerated Database Recovery.
One useful tool is the sys.dm_tran_persistent_version_store_stats dynamic management view. This view provides detailed statistics on the size and utilization of the version store for a specific database, as well as information about any background cleanup processes that may be running.
Additionally, SQL Server Management Studio provides a graphical view of the version store in the form of a new tab in the database properties window. This tab shows real-time statistics on the version store size and utilization, as well as the total number of versions and the oldest active transaction.
Limitations of SQL Server Accelerated Database Recovery
While Accelerated Database Recovery provides many benefits, there are also a few limitations to keep in mind.
Unsupported database features Not all database features are supported with Accelerated Database Recovery. For example, databases that use memory-optimized tables, table partitioning, or stretch database are not currently supported.
Increased disk space usage Accelerated Database Recovery can result in increased disk space usage due to the version store, which stores multiple versions of each modified page. This increased disk space usage may require additional planning and monitoring for large databases with high transactional volumes.
Potential performance impact In rare cases, Accelerated Database Recovery may cause a performance impact due to increased I/O operations required for log processing. However, this impact is typically minimal and is outweighed by the benefits of faster recovery times and improved availability.
Conclusion
SQL Server Accelerated Database Recovery provides a powerful new feature for improving database recovery times and reducing downtime. By leveraging innovative technology such as Log Sequence Number (LSN) filtering and persistent versioning, Accelerated Database Recovery enables faster and more reliable database recovery with minimal impact on workload.
While there are a few limitations to keep in mind, such as unsupported database features and increased disk space usage, the benefits of Accelerated Database Recovery far outweigh the potential drawbacks. If you’re running SQL Server 2019 Enterprise Edition or Azure SQL Database, consider enabling Accelerated Database Recovery to take advantage of its powerful benefits and improve your database’s availability and performance.
Disclaimer: The views expressed on this blog are mine alone and do not reflect the views of my company or anyone else. All postings on this blog are provided “AS IS” with no warranties, and confers no rights.
