opsmate ingest-prometheus-metrics-metadata
opsmate ingest-prometheus-metrics-metadata
ingests metrics metadata into the knowledge base.
OPTIONS
Usage: opsmate ingest-prometheus-metrics-metadata [OPTIONS]
Ingest prometheus metrics metadata into the knowledge base. The ingestion is
done via fetching the metrics metadata from the prometheus server, and then
storing it into the knowledge base. The ingested metrics metadata will be
used for providing context to the LLM when querying prometheus based metrics
Note this only enqueues the tasks to ingest metrics. To execute the actual
ingestion in the background, run `opsmate worker`. Please run: `opsmate
worker -w 1 -q lancedb-batch-ingest`
Options:
--prometheus-endpoint TEXT Prometheus endpoint. If not provided it uses
$PROMETHEUS_ENDPOINT environment variable,
or defaults to http://localhost:9090
[default: (dynamic)]
--prometheus-user-id TEXT Prometheus user id. If not provided it uses
$PROMETHEUS_USER_ID environment variable, or
defaults to empty string [default:
(dynamic)]
--prometheus-api-key TEXT Prometheus api key. If not provided it uses
$PROMETHEUS_API_KEY environment variable, or
defaults to empty string [default:
(dynamic)]
--loglevel TEXT Set loglevel (env: OPSMATE_LOGLEVEL)
[default: INFO]
--categorise BOOLEAN Whether to categorise the embeddings (env:
OPSMATE_CATEGORISE) [default: True]
--reranker-name TEXT The name of the reranker model (env:
OPSMATE_RERANKER_NAME) [default: ""]
--embedding-model-name TEXT The name of the embedding model (env:
OPSMATE_EMBEDDING_MODEL_NAME) [default:
text-embedding-ada-002]
--embedding-registry-name TEXT The name of the embedding registry (env:
OPSMATE_EMBEDDING_REGISTRY_NAME) [default:
openai]
--embeddings-db-path TEXT The path to the lance db. When s3:// is used
for AWS S3, az:// is used for Azure Blob
Storage, and gs:// is used for Google Cloud
Storage (env: OPSMATE_EMBEDDINGS_DB_PATH)
[default: /root/.opsmate/embeddings]
--contexts-dir TEXT Set contexts_dir (env: OPSMATE_CONTEXTS_DIR)
[default: /root/.opsmate/contexts]
--plugins-dir TEXT Set plugins_dir (env: OPSMATE_PLUGINS_DIR)
[default: /root/.opsmate/plugins]
--db-url TEXT Set db_url (env: OPSMATE_DB_URL) [default:
sqlite:////root/.opsmate/opsmate.db]
--auto-migrate BOOLEAN Automatically migrate the database to the
latest version [default: True]
--help Show this message and exit.
DESCRIPTION
The ingest-prometheus-metrics-metadata
command enqueues tasks to ingest prometheus-based metrics metadata into the Opsmate knowledge base. The ingested metrics metadata will be used for improve the retrieval and precision of the Prometheus query results generated by the LLM.
Note this only enqueues the tasks to ingest metrics. To execute the actual ingestion in the background, run opsmate worker
. Please run: opsmate worker -w 1 -q lancedb-batch-ingest
EXAMPLES
Ingest metrics metadata from Prometheus
The following command enqueues tasks to ingest metrics metadata from a Prometheus endpoint:
opsmate ingest-prometheus-metrics-metadata
By default, the command uses dynamic values for the Prometheus endpoint, user ID, and API key. You can specify these values explicitly as below.
opsmate ingest-prometheus-metrics-metadata \
--prometheus-endpoint="https://prometheus-prod-01-eu-west-0.grafana.net/api/prom" \
--prometheus-user-id="xxxxx" \
--prometheus-api-key="glc_xxxx"
If the endpoints are not provided, the command will use the values from the environment variables:
export PROMETHEUS_ENDPOINT="https://prometheus-prod-01-eu-west-0.grafana.net/api/prom"
export PROMETHEUS_USER_ID="xxxxx"
export PROMETHEUS_API_KEY="glc_xxxx"
opsmate ingest-prometheus-metrics-metadata
Running the worker to process the ingestion tasks
After enqueuing the tasks, you need to run a worker to process them:
opsmate worker -w 1 -q lancedb-batch-ingest