Longtime stop RDS

Question

**Description of the problem:**
1. We started stopping the Oracle RDS instance.
2. The RDS instance switched to the “Stopping” status.
3. Oracle database stopped successfully (Alertlog - Database Instance Shutdown Complete)
4. After 1.5 hours, the **RDS instance did not stop.**
5. A search at https://repost.aws/knowledge-center did not provide answers to our problem.
6. A Google search showed that such cases occur frequently.
7. The RDS was clearly “frozen” in this state and all we had to do was wait.
8. After 2.5 hours the instance stopped.

**Question:**
When RDS stops, does it always implicitly create a snapshot before stopping???

**Our assumptions:**
* The implicit snapshot is hidden from us and is performed to improve data security and availability.
* The time it takes to create an implicit snapshot depends on:
* number of I/O operations per second and
* number of changes since the last snapshot (snapshots are assumed to be incremental)
* database sizes

Answer

Hello @Max_K,

Question: When RDS stops, does it always implicitly create a snapshot before stopping???

Answer: Indeed, as described in [1], when you stop a DB instance, Amazon RDS preserves the metadata of the instance, such as its endpoint, DB parameter group, and option group membership. Additionally, Amazon RDS retains the transaction logs, enabling you to perform a point-in-time restore if required.

[1] https://docs.aws.amazon.com/cli/latest/reference/rds/stop-db-instance.html

Answer

Hello @Max_K,

When you stop a DB instance, Amazon RDS retains the DB instance's metadata, including its endpoint, DB parameter group, and option group membership. Amazon RDS also retains the transaction logs so you can do a point-in-time restore if necessary. While a database instance is stopped, Amazon RDS does not delete any of your automatic backups or transaction logs. This means you can do a point-in-time restore to any point within your specified automated backup retention window, even after an instance is re-started.

Here are a few key points on why it may be taking a long time to stop an RDS Oracle instance:

* Stopping an RDS instance entails a multi-step procedure, which may require a significant amount of time, especially for larger databases. This process involves tasks such as cache flushing, data writing to disk, and executing a safe shutdown of the database.
   * In scenarios where the RDS instance is part of a Multi-AZ deployment, the cessation process might be prolonged as it necessitates ensuring data synchronization between the primary and standby instances before the instance can be stopped.
   * The time it takes to stop an RDS instance can also depend on the size and complexity of the database, the number of active connections, and the overall load on the instance.

To help reduce the time it takes to stop an RDS Oracle instance, you can consider the following:

* Check for any long-running transactions or queries that might be delaying the stop process.
    * Disconnect active client connections before initiating the stop operation.
    * Whenever feasible, schedule the stop procedure during a maintenance window when the database load is minimal.
    * Evaluate the configuration and resource utilization of the RDS instance to ensure it is appropriately sized for your workload.

[+] https://docs.aws.amazon.com/cli/latest/reference/rds/stop-db-instance.html

Longtime stop RDS

Contenuto pertinente