I heard a frequent question at last week’s Percona Live conference regarding Percona Monitoring and Management (PMM): How much disk space should I allocate for PMM Server?
First, let’s review the three components of Percona Monitoring and Management that consume non-negligible disk space:
- Prometheus data source for the time series metrics
- Query Analytics (QAN) which uses Percona Server XtraDB (Percona’s enhanced version of the InnoDB storage engine)
- Orchestrator, also backed by Percona Server XtraDB
Of these, you’ll find that Prometheus is generally your largest consumer of disk space. Prometheus hits a steady state of disk utilization once you reach the defined storage.local.retention period. If you deploy Percona Monitoring and Management 1.1.3 (the latest stable version), you’ll be using a retention period of 30 days. “Steady state” in this case means you’re not adding or removing nodes frequently, since each node comes with its own 1k-7k metrics to be scraped. Prometheus stores a one-time series per metric scraped, and automatically trims chunks (like InnoDB pages) from the tail of the time series once they exceed the retention period (so the disk requirement per static list of metrics remains “fixed” for the retention period).
However, if you’re in a dynamic environment with nodes being added and removed frequently, or you’re on the extreme end like these guys who re-deploy data centers every day, steady state for Prometheus may remain an elusive goal. The guidance you find below may help you establish at least a minimum disk provisioning threshold.
QAN is based on a web application and uses Percona Server 5.7.17 as it’s datastore. The Percona QAN agent runs one instance per monitored MySQL server, and obtains queries from either the Slow log or Performance Schema. It performs analysis locally to generate a list of unique queries and their corresponding metrics: min, max, avg, med, and p95. There are dimensions based on Tmp table, InnoDB, Query time, Lock time, etc. Check the schema for a full listing, as there are actually 149 columns on this table (show create table pmm.query_class_metricsG). While the table is wide, it isn’t too long: PMM Demo is ~9mil rows and is approximately 1 row per distinct query per minute per host.
Finally, there is Orchestrator. While the disk requirements for Orchestrator are not zero, they are certainly dwarfed by Prometheus and QAN. As you’ll read below, Percona’s Orchestrator footprint is a measly ~250MB, which is a rounding error. I’d love to hear other experiences with Orchestrator and how large your InnoDB footprint is for a large or active cluster.
For comparison, here is the resource consumption from Percona’s PMM Demo site:
- ~47k time series
- 25 hosts, which is on average ~1,900 time series/host, some are +4k
- 8-day retention for metrics in Prometheus
- Prometheus data is ~40GB
- Which should not increase until we add more host, as this isn’t a dynamic Kubernetes environment 🙂
- QAN db is 6.5GB
- We don’t currently prune records, so this will continue to grow
- 90% of space consumed is in query_class_metrics, which is ~9mil rows
- Our first record is ~September 2016, but only in the past three months
- This is MySQL QAN only, the MongoDB nodes don’t write anything into QAN (yet… we’re working on QAN for MongoDB and hope to ship this quarter!!)
- Orchestrator db is ~250MB
- audit table is 97% of the space consumed, ~2mil rows
So back to the original question: How much space should I allocate for Percona Monitoring and Management Server? The favorite answer at Percona is “It Depends®,” and this case is no different. Using PMM Demo as our basis, 46GB / 25 hosts / 8 days = ~230MB/host/day or ~6.9GB/host/30 day retention period. For those of you running 50 instances in PMM, you should be provisioning ~400GB of disk.
Of course, your environment is likely to be different and directly related to what you do and don’t enable. For example, a fully verbose Percona Server 5.7.17 configuration file like this:
## PMM Enhanced options
with none of the mysqld_exporter features disabled:
can lead to an instance that has +4k metrics and will push you above 230MB/host/day. This is what the top ten metrics and hosts by time series count from the PMM Demo look like:
What does the future hold related to minimizing disk space consumption?
- The PMM development team is working on the ability to purge a node’s data without access to the instance
- Today you need to call pmm-admin purge from the instance – which becomes impossible if you’ve already terminated or decommissioned the instance!
- We are following Prometheus’ efforts on the 3rd Gen storage re-write in Prometheus 2.0, where InfluxDB will do more than just indices
- Again we are following Prometheus’ efforts on Remote Read / Remote Write so we can provide a longer-term storage model for users seeking > 30 days (another popular topic at PL2017)
- Allows us to store less granular data (every 5s vs. every 1s)
- Usage of Graphite, OpenTSDB, and InfluxDB as secondary data stores on the Remote end
I’d love to hear about your own experiences using Percona Monitoring and Management, and specifically the disk requirements you’ve faced! Please share them with us via the comments below, or feel free to drop me a line directly firstname.lastname@example.org. Thanks for reading!