We have successfully set it up with this configuration but it feels like a lot of boilerplate to be adding for people who would like to set this up.
extraContainers:
- name: relaychain-kvdb-introspector
image: paritytech/polkadot-introspector:438d3406
command: [
"polkadot-introspector",
"kvdb",
"--db",
"/data/chains/versi_v1_9/db/full",
"--db-type",
"rocksdb",
"prometheus",
"--port",
"9620"
]
resources:
limits:
memory: "1Gi"
ports:
- containerPort: 9620
name: relay-kvdb-prom
volumeMounts:
- mountPath: /data
name: chain-data
- name: parachain-kvdb-introspector
image: paritytech/polkadot-introspector:438d3406
command: [
"polkadot-introspector",
"kvdb",
"--db",
"/data/chains/versi_v1_9/db/full/parachains/db",
"--db-type",
"rocksdb",
"prometheus",
"--port",
"9621"
]
resources:
limits:
memory: "1Gi"
ports:
- containerPort: 9621
name: para-kvdb-prom
volumeMounts:
- mountPath: /data
name: chain-data
Note there should be the option to run 1 or 2 sidecars. 1 to monitor the main db and 1 for the parachain db (as an option for relay chains).
We also need to create the appropriate ServiceMonitor for loading data in Prometheus.
The polkadot-introspector kvdb tool can be used to monitor the database continuously. We should add support for running this exporter as a sidecar in the node helm chart.
We have successfully set it up with this configuration but it feels like a lot of boilerplate to be adding for people who would like to set this up.
Note there should be the option to run 1 or 2 sidecars. 1 to monitor the main db and 1 for the parachain db (as an option for relay chains).
We also need to create the appropriate ServiceMonitor for loading data in Prometheus.