Skip to content
By using AWS re:Post, you agree to the AWS re:Post Terms of Use
AWS re:Postre:Post
About re:Post

Question about RDS event subscription

0

Hi re:Post!
We have set up a topic, subscription, and RDS event subscription.

The SNS notification (email) is for all RDS instances in us-east-2 regarding a specific event category, "Failure".

q1) Does this apply to backup failures or instance failures?

q2) How do we ensure we notification for backup failures/errors/issues (ie: RDS-EVENT-0002) beyond completed normally?

The RDS event subscription is named, "RDS-Backup-Watch-Event-Subscription"

The SNS topic is called, "RDS_Backup_Watch"

The topic subscription ARN is: "ARN arn:aws:sns:us-east-2:910286192445:RDS_Backup_Watch:4c03ac80-db09-4fe3-82e4-e46993c6f195"

Thank you for your time and help!

Best Regards,

Donald

Topics
Application IntegrationDatabase
Tags
Amazon Relational Database ServiceAmazon Simple Notification Service (SNS)
Language
English

asked a year ago241 views

2 Answers
1
Accepted Answer

To answer your questions about RDS event subscriptions:

q1) When you subscribe to the "Failure" event category for RDS instances, this covers a broad range of failure events including both backup failures and instance failures. The "Failure" category encompasses various critical issues that can occur with your RDS instances.

q2) To ensure you receive notifications for backup failures/errors/issues (like RDS-EVENT-0002), you should specifically subscribe to the "backup" event category in addition to the "failure" category. This will ensure you receive comprehensive notifications about backup-related events.

Based on your current setup with "RDS-Backup-Watch-Event-Subscription" and the "RDS_Backup_Watch" SNS topic, I recommend modifying your event subscription to include both the "backup" and "failure" event categories. This can be done using the AWS Management Console or the AWS CLI.

For example, using the AWS CLI, you could update your subscription with a command like:

aws rds create-event-subscription \
--subscription-name RDS-Backup-Watch-Event-Subscription \
--source-type db-instance \
--event-categories '["backup","failure"]' \
--sns-topic-arn arn:aws:sns:us-east-2:910286192445:RDS_Backup_Watch

This will ensure you receive notifications for both backup failures and successful backup completions, giving you comprehensive monitoring of your backup processes.
Sources
Overview of Amazon RDS event notification - Amazon Relational Database Service
Troubleshooting DB issues for Amazon RDS Custom for Oracle - Amazon Relational Database Service
Troubleshooting DB issues for Amazon RDS Custom for SQL Server - Amazon Relational Database Service

answered a year ago

  • DC
    a year ago

    Excellent, thank you. We have added "Failure" in addition to "Backup" for the event categories.

0

Hello,

According to the AWS documentation, when you subscribe to the "Failure" event category for RDS instances, this covers a broad range of failure events including:

• Instance failures like incompatible network • Version Upgrade failures • Storage scale/ autoscaling • Other critical system failures

The "Failure" category is comprehensive and encompasses various types of critical issues that can occur with your RDS instances.

[+] Amazon RDS event categories and event messages - DB instance events - https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/USER_Events.Messages.html#USER_Events.Messages.instance

Q2: How do we ensure notification for backup failures/errors/issues (like RDS-EVENT-0002) beyond completed normally?

Since RDS doesn't generate explicit backup failure events, you should monitor the backup lifecycle pattern by subscribing to the "backup" category, which includes:

• RDS-EVENT-0001: Automated backup started • RDS-EVENT-0002: Automated backup completed • RDS-EVENT-0003: Manual backup started • RDS-EVENT-0004: Manual backup completed

Detection Logic:

  1. Normal Pattern: You should receive both "started" and "completed" events for each backup operation
  2. Failure Indicator: If you receive a "started" event but no corresponding "completed" event within the expected timeframe, this indicates a backup failure
  3. Missing Backup: If you don't receive any backup events during scheduled backup windows, this could indicate backup scheduling issues

I hope this helps.

AWS

answered a year ago

  • DC
    a year ago

    Thanks Sachin for the explanations and further input! - turns out we are using that same detection logic - we get 96 notifications of 48 RDS started and completed backups. If we have anything but 96 then we'll assume something is off.
    It is, however, incredible and shocking to hear that " RDS doesn't generate explicit backup failure events". In my 25 years as a IT professional and former Oracle DBA I find that condition unacceptable and a severe short-coming/non-enterprise level of database system design and programming. Wow. Best Regards, Donald

  • Sachin_M
    a year ago

    Hi Donald,

    You raise a valid point about backup failure notifications, and we completely understand your concerns about monitoring these critical operations. We do have a feature request in place to have a backup failure notifications. While the current workaround of monitoring the expected number of start/complete events (96 notifications for 48 RDS instances in your case) provides a functional solution, we recognize there's room for improvement in making this more straightforward and explicit. We appreciate you taking the time to share your feedback.

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Relevant content