Skip to main content

DB extractor metrics

The DB extractor can send performance metrics for remote monitoring and debugging.

Prometheus metrics

This section contains the full list of available metrics from the extractor. For information on the different types of metrics, see the Prometheus documentation. In general, Counter starts at zero and increases during a run of the extractor, while Gauge may change freely.

NameTypePurpose
db_extractor_start_timeGaugeThe timestamp (seconds) of when the extractor last started.
db_extractor_finish_timeGaugeThe timestamp (seconds) of when the extractor last finished.
db_extractor_memory_bytesGaugeThe memory usage of extractor in bytes.
db_extractor_cpu_percentTotal The CPU usage of extractor (0 - 100). Notice that this value is summed across all threads, so the value might exceed 100.
db_extractor_num_threadsGaugeThe number of threads used by the extractor.
db_extractor_rows_fetchedCounterThe number of rows fetched from a database, labeled on each query.
db_extractor_rows_uploadedCounterThe total number of rows successfully uploaded to CDF.
db_extractor_queries_runCounterThe number of successful queries.
db_extractor_queries_failedCounterThe number failed queries.
db_extractor_connection_numCounterThe number of concurrent database connections active. This is the number of active queries.
db_extractor_server_numCounterThe number of databases specified in configuration.
db_extractor_runningCounterThe state of extractor as a numeric value. 1 indicates running, 0 indicates not running.
db_extractor_infoInformation dictionary about the given extractor. This is the extractor's name and version.