typesense_exporter

README

Typesense Exporter

Prometheus exporter for various metrics about Typesense, written in Go, and just barely glued together to be functional.

Cluster metrics and API stats are exposed based on what Typesense itself provides via its endpoints. See their documentation for more information.

Installation

Check the packages or releases pages for pre-built options.

Docker

docker pull ghcr.io/scraton/typesense_exporter:main
docker run --rm -p 9115:9115 -e TYPESENSE_API_KEY=xyz ghcr.io/scraton/typesense_exporter:main

Configuration

NOTE: Just like the Elasticsearch Exporter, this exporter fetches information from a Typesense cluster on every scrape. Having too short a scrape interval could cause performance issues, so please plan accordingly.

Argument	Env Variable	Description	Default
listen-address	LISTEN_ADDRESS	address to listen on for metrics interface	:9115
telemetry-path	TELEMETRY_PATH	path under which to expose metrics	/metrics
typesense-url	TYPESENSE_URL	HTTP API address for Typesense node	http://localhost:8108
typesense-timeout	TYPESENSE_TIMEOUT	timeout for trying to get Typesense metrics	5s
typesense-api-key	TYPESENSE_API_KEY	API key for typesense
log-level	LOG_LEVEL	sets log level	info

Metrics

Please see Typesense's documentation for cluster metrics and API stats.

Name	Type	Cardinality	Help
typesense_api_stats_delete_latency_seconds	gauge	1	Latency for delete requests in seconds
typesense_api_stats_delete_requests_per_second	gauge	1	Requests per second for deletions
typesense_api_stats_import_latency_seconds	gauge	1	Latency for delete requests in seconds
typesense_api_stats_import_requests_per_second	gauge	1	Requests per second for imports
typesense_api_stats_json_parse_failures	counter	0	Number of errors while parsing JSON
typesense_api_stats_latency_seconds	gauge	3	Latency for each method and endpoint
typesense_api_stats_pending_write_batches	gauge	1	Pending write batches
typesense_api_stats_requests_per_second	gauge	3	Requests per second for each method and endpoint
typesense_api_stats_search_latency_seconds	gauge	1	Latency for search requests in seconds
typesense_api_stats_search_requests_per_second	gauge	1	Requests per second for searches
typesense_api_stats_total_requests_per_second	gauge	1	Requests per second for all endpoints
typesense_api_stats_total_scrapes	counter	0	Current total Typesense API stats scrapes
typesense_api_stats_up	gauge	0	Was the last scrape of the Typesense stats.json endpoint successful
typesense_api_stats_write_latency_seconds	gauge	1	Latency for write requests
typesense_api_stats_write_requests_per_second	gauge	1	Requets per second for writes
typesense_cluster_metrics_json_parse_failures	counter	0	Number of errors while parsing JSON
typesense_cluster_metrics_memory_active_bytes	gauge	1	Total active memory in use by Typesense
typesense_cluster_metrics_memory_allocated_bytes	gauge	1	Total allocated memory in use by Typesense
typesense_cluster_metrics_memory_fragmentation_ratio	gauge	1	Fragmentation ratio for Typesense memory
typesense_cluster_metrics_memory_mapped_bytes	gauge	1	Total mapped memory in use by Typesense
typesense_cluster_metrics_memory_metadata_bytes	gauge	1	Total memory used for metadata by Typesense
typesense_cluster_metrics_memory_resident_bytes	gauge	1	Total resident memory in use by Typesense
typesense_cluster_metrics_memory_retained_bytes	gauge	1	Total retained memory in use by Typesense
typesense_cluster_metrics_total_scrapes	counter	0	Current total Typesense cluster metrics scrapes
typesense_cluster_metrics_up	gauge	0	Was the last scrape of the Typesense metrics.json endpoint successful

Credit & License

Code is based on the original work done by elasticsearch_exporter.

Contributing & Development

Contributions are welcome. Please fork the project on GitHub and open Pull Requests for any proposed changes.

Building

# ensure promu utility is installed (one time step):
make promu 

promu crossbuild
make docker