PuppetDB has it's own dashboard, designed to show what's going on in the server. It is primarily aimed at making sense of the data that PuppetDB stores. It covers some performance metrics, like the JVM Heap, and also a quick active and inactive node count. The following information is available on PuppetDB:
- JVM Heap: Total memory heap size of database
- Active and inactive nodes: Nodes with information inside of PuppetDB
- Resources: Total resources seen in PuppetDB
- Resource Duplication: Total resources that are a duplicate that PuppetDB can serve (higher is better)
- Catalog Duplication: Total catalogs that are a duplicate that PuppetDB can serve (higher is better)
- Command Queue: Number of commands waiting to be run
- Command Processing: How long commands take to execute against the database
- Processed: Number of queries processed since startup
- Retried: Number of queries that had to be run more than once
- Discard: Number of queries that did not return a value
- Rejected: Number of queries that were rejected
- Enqueuing: Average amount of time spent waiting to write to the database
- Command Persistence: The time it takes to move data from memory to disk
- Collection Queries: Collection query service time in seconds
- DB Compaction: Round trip time for database compaction
- DLO Size on Disk: Dynamic large object size on disk
- Discarded Messages: Messages that did not enter PuppetDB
- Sync Duration: Amount of time it takes to sync data between databases
- Last Synced: How many seconds since the last database sync
We can use this information to check the status of our PuppetDB server. We want to see a high resource duplication and catalog duplication, which speeds up our overall runs of Puppet using PuppetDB. Our JVM heap can let us know how we're doing on memory usage. Active and inactive nodes help us understand what's being stored in PuppetDB, and what is on it's way out. Most other data is metrics surrounding the database itself, letting us know the health of the PostgreSQL server. Once we understand some simple live metrics, we can start looking at tuning our environment.