Performance

Review multi-level performance graphs within Middle

Middle provides performance graphs on the instance, account, and workflow levels. This allows you to quickly monitor what's happening in your Middle account to ensure performance.

Graph types

There are six bar graphs for your review:

  • Script executions - all scripts being performed

  • Script executions (syncs) - all scripts being performed to sync data into Middle or a given account

  • Script executions (actions) - all scripts being performed to execute actions

  • Syncs - statistics about data being synced into Middle or a given account

  • Activity triggers - statistics about activity-based workflow triggers executing

  • Workflows - statistics about workflow performance

Adjusting graph range and period

The range for each graph can be adjusted from one hour to four weeks. In addition, you can select the time period for each bar on the graph, ranging from five minutes to one week.

Available statistics

Each graph provides a number of statistics. You have the option of toggling these on and off to hone in on the data you're most interested in. As you hover over the bar graph you can see a numerical breakdown for each statistic in that given period. Depending on the graph, the available statistics can vary.

For script execution graphs

  • Middle system error - Something went wrong with Middle's platform that prevented the script's results from being processed

  • Developer fatal error - App-defined code raised an exception, timed out, or ran out of memory

  • Developer non-fatal error - App-defined code raised a special exception saying that the user needs to change their input (right now, only an AuthorizationError in OAuth flow options will show up as a non-fatal error)

  • Restart invocation - The script execution failed and was restarted as a result

  • Successful exit - The script execution succeeded with no failures

  • Execution time - The amount of time the script executions in the defined period took to execute

  • Bytes returned (kb) - The amount of data returned by script executions in kilobytes

For syncs

  • Invalid records - Records synced that did not pass validation

  • Duplicate records - When Middle syncs a record that has already been seen and stored

  • Stored records - Total number of records stored by syncs in the given period

  • Updated records - Total number of records that were updated in the given period

  • Bytes processed (kb) - Binary kilobytes processed by syncs in the given period

  • Bytes stored (kb) - Binary kilobytes stored by syncs in the given period

For activity triggers

  • Count of system error - Something went wrong with Middle's platform that prevented the script's results from being processed

  • Count of errors ("TooManyRestarts" or "TimeLimitExceeded") - Activity triggers are only tried a certain number of times and for a certain wall clock duration before we stop trying, configurable by the user in the Trigger editor

  • Count of passes - Triggers that were successfully executed

  • Count of discards - Number of workflows that were discarded before being run by trigger discard logic

  • Count of requeues - Number of workflows that were requeued due to the requeue logic in the trigger being met (likely due to related records not being synced)

  • Node parameter: Execution time (in seconds) - The amount of time taken to execute triggers in the given time period

  • Node parameter: Sum of records pulled from db - The number of records in the Middle database that were referenced by triggers in the given time period

  • Node parameter: Bytes of records pulled from db (kb) - The number of kilobytes in the Middle database referenced by triggers in the given time period

For workflows

  • Count OK - Number of workflows that were successfully executed

  • Count Fail - Number of workflows that ended with a failure

  • Workflow: Execution time (in seconds) - How long the workflow took to execute, not counting breaks to run actions (i.e., processor time only)

  • Node parameter: Execution time (in seconds) - How long each Node Parameter took to execute. Big numbers here may indicate large deeply nested lookups (like looking at all check-ins for all users in all locations) or other kinds of processing.

  • Node parameter: Sum of records pulled from db - The number of records in the Middle database that were referenced by workflows in the given time period

  • Node parameter: Bytes of records pulled from db (kb) - The number of binary kilobytes in the Middle database referenced by workflows in the given time period

Last updated