Pipeline Metrics
This reference lists all of the metrics that Feldera exports through
its /metrics endpoint in Prometheus exposition format. It is
automatically generated using the documentation embedded in Prometheus
output.
All of the metrics exported by a particular Feldera pipeline are
labeled with the pipeline's UUID as pipeline and its name as
pipeline_name. Some metrics have additional labels, as documented
below.
See Monitoring and Profiling for a guide to setting up Prometheus and Grafana with Feldera. The Feldera template dashboard is a sample Grafana dashboard for Feldera.
Process Metrics
These metrics report statistics for a running Feldera pipeline process. When a pipeline process is killed and restarts from a checkpoint, the new process's metrics are for it alone, not cumulative with any previous instantiations.
These metrics are intended to match the standard Prometheus definitions.
| Name | Type | Description |
|---|---|---|
process_cpu_seconds_total | counter | Total user and system CPU time spent in seconds. |
process_max_fds | gauge | Maximum number of open file descriptors. |
process_open_fds | gauge | Number of open file descriptors. |
process_resident_memory_bytes | gauge | Resident set size in bytes. |
process_start_time_seconds | counter | Start time of the process in seconds since the Unix epoch. |
process_threads | gauge | Number of OS threads in the process. |
process_virtual_memory_bytes | gauge | Virtual memory size in bytes. |
process_virtual_memory_max_bytes | gauge | Maximum amount of virtual memory available in bytes. |
Feldera metrics
These metrics report statistics for Feldera operations.
| Name | Type | Description |
|---|---|---|
feldera_checkpoint_delay_seconds | histogram | Sub-duration of feldera_checkpoint_runtime_seconds during which pipeline execution was blocked. |
feldera_checkpoint_records_processed_total | counter | Total number of records that had processed when the most recent checkpoint successfully committed. |
feldera_checkpoint_runtime_seconds | histogram | Time to run checkpoint operations, in seconds, including time that the pipeline could continue executing along with the checkpoint. |
feldera_checkpoint_written_bytes | histogram | Amount of data written to storage during checkpoints, in bytes. |
DBSP metrics
These metrics report statistics for DBSP, the low-level mechanism on which Feldera is built.
| Name | Type | Description |
|---|---|---|
compaction_stall_duration_seconds | counter | Time in seconds a worker was stalled waiting for more merges to complete. |
dbsp_operator_checkpoint_latency_seconds | histogram | The time that individual operator checkpoint operations delayed the pipeline, in seconds. (Because checkpoints run in parallel across workers, these will add up to more than feldera_checkpoint_delay_seconds.) |
dbsp_runtime_elapsed_seconds_total | counter | Time elapsed while the pipeline is executing a step, multiplied by the number of foreground and background threads, in seconds. |
dbsp_step_latency_seconds | histogram | Latency of DBSP steps over the last 60 seconds or 1000 steps, whichever is less, in seconds |
dbsp_steps_total | counter | Total number of DBSP steps executed. |
output_stall_seconds | gauge | If the pipeline is currently stalled because one or more output connectors' output buffers were full, this is the time in seconds for which it has been stalled. If the pipeline is not currently stalled, this is zero. If this is nonzero, then the output connectors causing the stall can be identified by observing which values of output_connector_queued_records are greater than or equal to the configured maximum (which defaults to 1,000,000). |
output_stall_seconds_total | counter | Time in seconds that the pipeline was stalled because one or more output connectors' output buffers were full. This value is greater than or equal to output_stall_seconds. |
Record Processing
These metrics report overall counts of records as they pass through the pipeline. They accumulate across checkpoint and resume.
| Name | Type | Description |
|---|---|---|
output_buffered_batches | gauge | Number of batches of records currently buffered by the output connector. |
records_input_buffered | gauge | Total amount of data currently buffered by all endpoints, in records. |
records_input_buffered_bytes | gauge | Total amount of data currently buffered by all endpoints, in bytes. |
records_input_bytes_total | counter | Total amount of data received from all connectors, in bytes. |
records_input_total | counter | Total amount of data received from all connectors, in records. |
records_late_total | counter | Number of records dropped due to LATENESS annotations. |
records_processed_bytes_total | counter | Total amount of input processed by the pipeline, in bytes. |
records_processed_total | counter | Total amount of input processed by the pipeline, in records. |
Storage Performance
These metrics report the performance of storage, which allows Feldera to work with data larger than memory.
| Name | Type | Description |
|---|---|---|
files_created_total | counter | Total number of files created. |
files_deleted_total | counter | Total number of files deleted. |
storage_byte_seconds_total | counter | Storage usage integrated over time during this run of the pipeline, in bytes × seconds. |
storage_cache_usage_bytes | gauge | The number of bytes of memory currently in use for caching data on storage. |
storage_cache_usage_limit_bytes_total | counter | The limit for the number of bytes of memory for caching data on storage. |
storage_read_block_bytes | histogram | Sizes in bytes of blocks read from storage. |
storage_read_latency_seconds | histogram | Read latency for storage blocks in seconds |
storage_sync_latency_seconds | histogram | Sync latency in seconds |
storage_usage_bytes | gauge | The number of bytes of storage currently in use |
storage_write_block_bytes | histogram | Sizes in bytes of blocks written to storage. |
storage_write_latency_seconds | histogram | Write latency for storage blocks in seconds |
Pipeline Status
These metrics report the status of the pipeline.
| Name | Type | Description |
|---|---|---|
pipeline_complete | counter | Transitions from 0 to 1 when pipeline completes. |
pipeline_start_time_seconds | counter | Start time of the pipeline in seconds since the Unix epoch. This will be earlier than process_start_time_seconds if the pipeline resumed from a checkpoint. This will be zero if the pipeline resumed from a checkpoint produced by a pipeline too old to record its start time. |
Input Connectors
These metrics are per-input connector, labeled with endpoint set to
the name of the input connector, which is either the name assigned in
the SQL program or automatically generated as unnamed-<number>,
where <number> counts starting from 1 for the first connector for a
given table.
These metrics accumulate across checkpoint and resume.
For byte counters, for some input connectors, such as columnar formats, bytes are difficult to attribute accurately to records, so Feldera approximates. Feldera also approximately attributes byte counts to records when it processes only some of the records in a batch in a DBSP step. This approximation is corrected when the remainder of the batch is processed in a subsequent step, so it is invisible to users unless a pause or checkpoint happens mid-batch.
| Name | Type | Description |
|---|---|---|
input_connector_barrier | gauge | Whether the input connector is currently a barrier for checkpointing/suspend (1 for true, 0 for false). |
input_connector_buffered_records | gauge | Amount of data currently buffered by an input connector, in records. |
input_connector_buffered_records_bytes | gauge | Amount of data currently buffered by an input connector, in bytes. |
input_connector_bytes_total | counter | Total number of bytes received by an input connector. |
input_connector_completion_latency_seconds | histogram | Time between when the connector receives new data and when the pipeline processes this data, computes output updates, and sends these updates to all output connectors, over the last 600 seconds or 10,000 samples. |
input_connector_delta_phase | gauge | Current phase: 0=loading_snapshot, 1=follow/streaming, 2=completed. |
input_connector_delta_snapshot_completed_seconds | gauge | Unix epoch seconds when the snapshot phase finished (0 if not yet complete). |
input_connector_delta_snapshot_records_total | counter | Total records loaded during the snapshot phase. |
input_connector_end_of_input | gauge | Whether the input connector has reached end of input (1 for true, 0 for false). |
input_connector_errors_parse_total | counter | Total number of errors encountered parsing records received by the input connector. |
input_connector_errors_transport_total | counter | Total number of errors encountered by the input connector at the transport layer. |
input_connector_extra_memory_bytes | gauge | Additional memory used by an input connector beyond that used for buffered records. |
input_connector_processing_latency_seconds | histogram | Time between when the connector receives new data and when the pipeline processes this data and computes output updates, over the last 600 seconds or 10,000 samples. |
input_connector_records_total | counter | Total number of records received by an input connector. |
input_connector_running | gauge | Whether the input connector is running (1) or paused by the user (0). |
Output Connectors
These metrics are per-output connector, labeled with endpoint set to
the name of the output connector, which is either the name assigned in
the SQL program or automatically generated as unnamed-<number>,
where <number> counts starting from 1 for the first connector for a
given view.
These metrics accumulate across checkpoint and resume.
| Name | Type | Description |
|---|---|---|
output_connector_buffered_records | gauge | Number of records currently buffered by the output connector. |
output_connector_bytes_total | counter | Total number of bytes of records sent by the output connector. |
output_connector_errors_encode_total | counter | Total number of errors encountered encoding records to send. |
output_connector_errors_transport_total | counter | Total number of errors encountered at the transport layer sending records. |
output_connector_extra_memory_bytes | gauge | Additional memory used by an output connector beyond that used for buffered records. |
output_connector_queued_batches | gauge | Number of batches of records currently queued by the output connector. |
output_connector_queued_records | gauge | Number of records currently queued by the output connector. |
output_connector_records_total | counter | Total number of records sent by the output connector. |
Checkpoint Synchronization
These metrics report the status of checkpoint synchronization.
| Name | Type | Description |
|---|---|---|
checkpoint_sync_pull_duration_seconds | histogram | Time taken to pull a checkpoint from object store in seconds. |
checkpoint_sync_pull_failures | counter | Number of failures when pulling a checkpoint. |
checkpoint_sync_pull_success | counter | Number of checkpoints pulled successfully. |
checkpoint_sync_pull_transfer_speed_bytes_per_second | histogram | Transfer speed when pulling a checkpoint, in bytes per second. |
checkpoint_sync_pull_transferred_bytes | histogram | Bytes transferred when pulling a checkpoint. |
checkpoint_sync_push_duration_seconds | histogram | Time taken to push a checkpoint to object store in seconds. |
checkpoint_sync_push_failures | counter | Number of failures when pushing a checkpoint. |
checkpoint_sync_push_success | counter | Number of checkpoints pushed successfully. |
checkpoint_sync_push_transfer_speed_bytes_per_second | histogram | Transfer speed when pushing a checkpoint, in bytes per second. |
checkpoint_sync_push_transferred_bytes | histogram | Bytes transferred when pushing a checkpoint. |
Transactions
These metrics report the status of transactions.
| Name | Type | Description |
|---|---|---|
transaction_commit_seconds | histogram | Transaction commit time, that is, from starting commit to finishing commit. |
transaction_completed_operators | gauge | Number of operators that have been fully flushed while the current transaction is committing. This is 0 if no transaction is active, or if a transaction is running but has not yet started committing. |
transaction_in_progress_operators | gauge | Number of operators that are currently being flushed while the current transaction is committing. This is 0 if no transaction is active, or if a transaction is running but has not yet started committing. |
transaction_ingest_seconds | histogram | Transaction ingestion time, that is, from transaction start to start of commit. |
transaction_remaining_operators | gauge | Number of operators that have not started flushing while the current transaction is committing. This is 0 if no transaction is active, or if a transaction is running but has not yet started committing. |
transaction_state | gauge | 0 when no transaction is active, 1 when a transaction has started, 2 while a transaction is committing. |