github.com/authzed/spicedb@v1.32.1-0.20240520085336-ebda56537386/TELEMETRY.md (about)

     1  # SpiceDB Telemetry
     2  
     3  SpiceDB reports metrics that are used to understand how clusters are being configured and the performance they are experiencing.
     4  The intent of collecting this information is to prioritize development that will have the most impact on the community.
     5  
     6  Telemetry can always be disabled by providing the flag `--telemetry-endpoint=""`.
     7  
     8  ## Implementation
     9  
    10  Telemetry is reported via the Prometheus Remote Write protocol.
    11  Any metrics prefixed with `spicedb_telemetry` are reported hourly to `telemetry.authzed.com`.
    12  You can find all of the code in [internal/telemetry][telemetry-package].
    13  
    14  [telemetry-package]: https://github.com/authzed/spicedb/tree/main/internal/telemetry
    15  
    16  ## Collected metrics
    17  
    18  ### spicedb_telemetry_info (Gauge)
    19  
    20  Information about the SpiceDB environment.
    21  
    22  - CPU Architecture: architecture of the CPU
    23  - vCPUs: number of CPUs reported by the Go runtime
    24  - OS: operating system family as reported by the Go Runtime
    25  - Datastore Engine: datastore engine configured
    26  - Cluster ID: unique identifier for a cluster's datastore
    27  - NodeID: unique identifier for the node, usually the hostname
    28  
    29  ### spicedb_telemetry_object_definitions_total (Gauge)
    30  
    31  Count of the number of objects defined by the schema.
    32  
    33  - Cluster ID: unique identifier for a cluster's datastore
    34  - NodeID: unique identifier for the node, usually the hostname
    35  
    36  ### spicedb_telemetry_relationships_estimate_total (Gauge)
    37  
    38  Rough estimate of the number of relationships stored in the datastore.
    39  
    40  - Cluster ID: unique identifier for a cluster's datastore
    41  - NodeID: unique identifier for the node, usually the hostname
    42  
    43  ### spicedb_telemetry_dispatches (Histogram)
    44  
    45  Histogram of cluster dispatches performed by the instance.
    46  
    47  - Method: gRPC method of the dispatch request
    48  - Cached: boolean of whether or not the dispatch was a cache hit
    49  - Cluster ID: unique identifier for a cluster's datastore
    50  - NodeID: unique identifier for the node, usually the hostname