Databases offer a lot of metrics. How does the enterprise go about wading through it all to uncover the insights businesses and customers are looking for?
DBTA recently held a webinar with Katy Farmer, developer advocate, InfluxData, who discussed which metrics can offer insight into the database and why they matter.
Businesses monitor their databases to detect outages, minimize downtime, for storage requirements for predictive analysis, to make debugging easier, and system health, according to Farmer.
However, the problem is, databases collect everything and makes it more difficult to sift through the metrics that matter.
Metrics that matter include:
- Throughput: How much is the database doing?
- Queries
- Transactions
- Points/sec
- Bytes/sec
- Execution time: How long does it take the database to do its job?
- Monitoring execution time is not the same as improving execution time
- Concurrency: How many jobs is the database doing at once?
- Managed by concurrency algorithms
- Help identify trends
- Build a bigger picture of the database
- Utilization: What percentage of the time was the database available?
- What are the consequences of your database being unavailable?
InfluxData can help discover what’s behind these meaningful metrics by offering a solution that acts as a purpose built time-series database, according to Farmer.
An archived on-demand replay of this webinar is available here.