Mar 30, 2022
Keeping your database “Otterly” healthy
Bohan Zhang | Ruogu Du
It’s no joke if your database is sick. Unexpected downtime or sluggish performance is bad news for business and will lead co-workers to turn on each other. That’s why you need to groom your database like an otter to keep it running smoothly, stay available, and avoid slow performance.
Carving time out for regularly scheduled grooming is easier said than done when you have so many competing priorities to keep your database and your core business humming along. To make matters worse, database vendors now update their software on the regular; making it hard to keep up with every product version.
The good news? We’re helping you keep your database in top shape with OtterTune Database Health Checks. This new feature automatically detects and informs you of possible issues in your database in real-time and provides recommendations to fix what’s going on. As with all things OtterTune, we’re here to help you keep your database optimized, efficient, scalable and highly available.
What database health automatically checks
Using Database Health Checks, you can stay on top of everything from your buffer cache hit rate to your disk space usage or the number of joins without using indexes. The Database Health checks are unique to different target databases, so the number of possible checks varies a little depending on what database you’re running.
Database Health also provides in-depth descriptions of each health check:
OtterTune checks whether your database resource utilization is within a healthy range. For example, if your disk IOPS is under-utilized , it will recommend reducing provisioned IOPS to save cost. If it’s over-utilized, it will suggest increasing the provisioned IOPS.
Database Connection Utilization
Disk IOPS Utilization
Disk Space Usage
Local Storage Usage
OtterTune checks whether your database cache management is efficient. For example, if your buffer cache hit rate is low, it will recommend increasing your database’s buffer pool size.
Buffer Cache Hit Rate
Index Cache Hit Rate
Opened Table Rate
Sort Merge Passes Ratio
OtterTune checks whether your database index, autovacuum, and checkpoint management are efficient. For example, if your database has many joins without using indexes, it will recommend adding missing indexes and optimizing slow queries.
Dead Tuple Ratio
Index Scan Ratio
Join Without Indexes Rate
Requested Checkpoint Ratio
You can get started quickly with Database Health Checks. The first time you load the dashboard, it usually takes 20 to 30 seconds. After that, the cache speeds up the loading process.
Where Database Health gets its data
Some of OtterTune’s Database Health Checks use data from Amazon CloudWatch and RDS Performance Insights. Other checks require data from the OtterTune Agent. Rest assured that Database Health, like the OtterTune Agent, doesn’t need access to schema, indexes, queries, or application data. And you don’t need to expose your database for external access.
So many clams, so little time
We otters need to eat 25 percent of our body weight in food every day. That takes a lot of time and effort. Let OtterTune Database Health do the hunting, digging and cracking for you.