Mastering Synchronous Data Guard: Advanced Configurations and Pitfall Avoidance
Data Guard is great. I love Data Guard. But more often than not I see it being misunderstood and sometimes misconfigured.
This session goes beyond standard Data Guard implementations, empowering attendees to harness the full potential of synchronous replication. We will explore critical, often-overlooked configurations and advanced features essential for building truly resilient and performant disaster recovery solutions.
In my talk, we'll cover the following:
- The topology - what to consider and where to place your observers, lessons learned (what you should NOT do) and some less obvious tips.
- How to plan your deployments of DG, and how to test them correctly to avoid surprises.
- SYNC/(NO)AFFIRM - what to look out for when you decide to go full SYNC, and how to correctly measure its impact on your system (related wait events, rfs traces with examples).
DEMO: We'll go through a real example of a spotty network and identifying the issue using ASH/AWR metrics and a follow-up of going through the RFS traces.
- Split-brain - and how to address it correctly. Letting your primary database shut down automatically is scary. We'll go over setting up the minimum required environment to keep things safe and the most common pitfalls. This is a balancing act of application uptime requirements and keeping your data consistent.
DEMO: I'll show some examples of what NOT to do with your FSFO settings. Seemingly correct and sensible settings which may have disastrous consequences.
- Benefits of using DataGuard together with RAC. What does it enable and how impactful are they? We'll glance over application continuity and FAN.
- Data Guard in Multitenant - what it can and cannot do on 19c (and what's improved in 21c and 23ai).
This session focuses on the more advanced topics. It is designed for DBAs and architects who are planning to use DG as their primary DR technology. It assumes basic knowledge of Data Guard and it will not go over the basics.
Monitoring Oracle DBs - the current tools landscape
What are some good options for the Oracle DB monitoring these days?
In this slightly opinionated LAB, I want to explore what are some good modern options to monitor your databases (or fleet of them). We'll consider different approaches to monitoring and play around with some of the best tools for the job.
We'll dive into:
- OEM 24ai - what's good actually about it, using emcli to automate most of it
- Grafana (with different backends and oracle's node_exporter) - some good practical use cases
- ELK stack - the king of log analysis for all your databases - I'll share the techniques I use to analyze Oracle's logs at large scale and other cool tricks for data analysis
- analyzing Oracle logs with local LLMs - a feasible approach
- my own monitoring framework - opinionated portable PL/SQL framework to measure, record and analyze metrics in databases you don't have direct access to
Online Schema Changes: Real-world Experience with Edition-Based Redefinition
Rolling deployments have become standard for application updates, but database schema changes remain a major challenge. Developers often resort to workarounds that compromise data quality or require intricate coding solutions.
What if there was a better way? Oracle Edition-Based Redefinition (EBR)—combined with Online Table Reorganization—can seamlessly evolve database schemas without downtime.
In this session, we'll start with the fundamentals of EBR before diving into a real-world case study with ČSOB, a major financial institution from the Czech Republic. You'll hear firsthand how they leveraged EBR to implement zero-downtime schema changes, ensuring business continuity while lowering deployment risks. We'll also demonstrate a live deployment to showcase EBR in action.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top