Request a Demo
53.3. Viewing Profiler Graphs
Prev UpChapter 53. ProfilerHome Next

53.3. Viewing Profiler Graphs #

Before performing this instruction:

To view profiler graphs:

  1. Go to profiler graphs in one of the following ways:

    • Using the monitoring section:

      1. In the navigation panel, go to Monitoring → Profiler.

      2. From Instance, select the instance.

    • Using the configuration section of an instance:

      1. In the navigation panel, go to Infrastructure → Instances.

      2. Click the name of the instance.

      3. In the navigation panel, go to Profiler.

  2. From Database, select the database.

  3. From Server, select the profiler server.

  4. (Optional) In Period, specify the time period.

    If you do not specify a value, the profiler graphs are displayed for the entire time range.

  5. Click Select.

The following profiler graphs will be displayed:

  • PostgreSql Instance: tuples: Operations with rows per second in the database. Available metrics:

    • Tuples returned: The number of live rows fetched by sequential scans and index entries returned by index scans.

      This metric corresponds to the tup_returned column of the pg_stat_database view.

    • Tuples fetched: The number of live rows fetched by index scans.

      This metric corresponds to the tup_fetched column of the pg_stat_database view.

    • Tuples inserted: The number of rows inserted by queries.

      This metric corresponds to the tup_inserted column of the pg_stat_database view.

    • Tuples updated: The number of rows updated by queries.

      This metric corresponds to the tup_updated column of the pg_stat_database view.

    • Tuples deleted: The number of rows deleted by queries.

      This metric corresponds to the tup_deleted column of the pg_stat_database view.

    For more information about these metrics, refer to the official Postgres Pro documentation on the pg_stat_database view.

  • PostgreSql bgwriter buffers: Operations with buffers per second in the database. Available metrics:

    • Checkpoints buffers written: The number of buffers written during checkpoints and restartpoints.

      This metric corresponds to the buffers_written column of the pg_stat_checkpointer view.

    • Background buffers written: The number of buffers written by the background writer.

      This metric corresponds to the buffers_clean column of the pg_stat_bgwriter view.

    • Backend buffers written: The number of buffers written directly by the backend.

      In Postgres Pro 16, this metric corresponds to the buffers_backend column of the pg_stat_bgwriter view. In Postgres Pro 17, this metric is collected by the pg_stat_io view.

    • Number of buffers allocated: The number of allocated buffers.

      This metric corresponds to the buffers_alloc column of the pg_stat_bgwriter view.

    For more information about these metrics, refer to the official Postgres Pro documentation on the pg_stat_bgwriter, pg_stat_io, and pg_stat_checkpointer views.

  • PostgreSql bgwriter write/sync: Synchronization and writing operations with buffers per second in the database. Available metrics:

    • Bgwriter interrupts: The number of times the background writer stopped a cleaning scan due to writing too many buffers.

      This metric corresponds to the maxwritten_clean column of the pg_stat_bgwriter view.

    • Backend fsync count: The number of times the backend executed its own fsync call. Normally, these calls are executed by the background writer, even when the backend performs its own writes.

      In Postgres Pro 16 or lower, this metric corresponds to the buffers_backend_fsync column of the pg_stat_bgwriter view. In Postgres Pro 17, this metric is collected by the pg_stat_io view.

    For more information about these metrics, refer to the official Postgres Pro documentation on pg_stat_bgwriter and pg_stat_io.

  • PostgreSql checkpoints count: Operations with checkpoints per second in the database. Available metrics:

    • Scheduled checkpoints: The number of scheduled checkpoints that were completed due to timeout. Scheduled checkpoints can be skipped if the server has been idle since the last checkpoint. Both completed and skipped scheduled checkpoints are counted.

      This metric corresponds to the num_timed column of the pg_stat_checkpointer view.

    • Requested checkpoints: The number of requested checkpoints that were completed.

      This metric corresponds to the num_requested column of the pg_stat_checkpointer view.

    For more information about these metrics, refer to the official Postgres Pro documentation on the pg_stat_checkpointer view.

  • PostgreSql checkpoints write/sync: The time per second spent writing and synchronizing files during checkpoints in the database. Available metrics:

    • Checkpoint write time (s): The time in seconds spent writing files to disk while completing checkpoints and restartpoints.

      This metric corresponds to the write_time column of the pg_stat_checkpointer view.

    • Checkpoint sync time (s): The time in seconds spent synchronizing files to disk while completing checkpoints and restartpoints.

      This metric corresponds to the sync_time column of the pg_stat_checkpointer view.

    For more information about these metrics, refer to the official Postgres Pro documentation on the pg_stat_checkpointer view.

  • PostgreSql Instance: events: Operations with events per second in the database. Available metrics:

    • Conflicts: The number of queries canceled due to conflicts with recovery. Conflicts can occur only on standby servers.

      This metric corresponds to the conflicts column of the pg_stat_database view.

      For more information about conflicts, refer to the official Postgres Pro documentation on the pg_stat_database_conflicts view.

    • Deadlocks: The number of deadlocks.

      This metric corresponds to the deadlocks column of the pg_stat_database view.

    • Rollbacks: The number of rolled back transactions.

      This metric corresponds to the xact_rollback column of the pg_stat_database view.

    • Commits: The number of committed transactions.

      This metric corresponds to the xact_commit column of the pg_stat_database view.

    For more information about these metrics, refer to the official Postgres Pro documentation on the pg_stat_database view.

  • PostgreSql: cache hit ratio: This graph displays the Cache hit ratio metric that provides the percentage of the data received from the buffer cache per second in the database.

    This metric is based on the blks_hit and blks_read columns of the pg_stat_database view and is calculated as follows:

    blks_hit / (blks_hit + blks_read)

    For more information about the blks_hit and blks_read columns, refer to the official Postgres Pro documentation on the pg_stat_database view.

  • PostgreSql temp: bytes written: This graph displays the Bytes written metric that provides the amount of data written to temporary files by queries per second in the database. All temporary files are counted, regardless of why they were created and of the log_temp_files parameter value.

    This metric corresponds to the temp_bytes column of the pg_stat_database view.

    For more information about this metric, refer to the official Postgres Pro documentation on the pg_stat_database view.

  • PostgreSql temp: files created: This graph displays the Number of files metric that provides the number of temporary files created by queries per second in the database. All temporary files are counted, regardless of why they were created (sorting, hashing) and of the log_temp_files parameter value.

    This metric corresponds to the temp_files column of the pg_stat_database view.

    For more information about this metric, refer to the official Postgres Pro documentation on the pg_stat_database view.

  • PostgreSql archive command: Archiving operations with WAL files per second in the database. Available metrics:

    • WAL segments archived: The number of successfully archived WAL files.

      This metric corresponds to the archived_count column of the pg_stat_archiver view.

    • WAL segments archive failed: The number of failed attempts to archive WAL files.

      This metric corresponds to the failed_count column of the pg_stat_archiver view.

    For more information about these metrics, refer to the official Postgres Pro documentation on the pg_stat_archiver view.

  • PostgreSQL WAL write speed: This graph displays the WAL generated metric that provides the amount of WAL in bytes generated per second in the database.

    This metric corresponds to the wal_bytes column of the pg_stat_wal view.

    For more information about this metric, refer to the official Postgres Pro documentation on the pg_stat_wal view.

You can perform the following actions using the icons in the top-right corner of the planner graphs:

  • To build a report using a graph, click Select the period for the report.

    For more information about building a report, refer to Section 53.5.1.

  • To reset the selected period when building a report, click Reset.

  • To download the graph in the PNG format, click Save as Image.

Prev Up Next
53.2. Samples Home 53.4. Sampling Schedules
epub pdf