Automation Services instance performance counters

When you monitor the instance of Automation Services, you can review the detailed information provided in the Performance Counters area to assess the performance measures of the service instance. All memory values are shown in MB.

To access the table, follow the instructions in Accessing performance data.

Note:

To reset the number of cached queries to external data sources, click Clear cache for all data connections.
To run a full garbage collection twice (to clear memory no in use), click Run a full GC(2). Remember that a full garbage collection may take time and the service will be unresponsive during the running.

For information about using performance counters, see Performance troubleshooting.


Performance Counter	Description
# .NET induced GC	The number of times that an induced GC has been performed. This is .NET Common Language Runtime (CLR) Memory.
% time in GC	The percentage of processor time spent in GC. This is .NET CLR Memory.
Accumulated processor time	The accumulated number of CPU seconds since the service start. If this number is consistently high, you might have performance problems related to CPU consumption. See Performance troubleshooting for more information.
Active threading jobs	The number of active jobs in graphical tables.
Active threads in thread pool	The number of active threads in thread pool (in .NET).
Available memory	The total MBytes available, based on standard performance counter in the category Memory. If this value is low compared to Total working memory, then you might have performance problems related to RAM. See Performance troubleshooting for more information.
Available memory %	The memory that is still available, as a percentage of the total.
Available temp disk space	The amount of available disk space allocated as temporary.
Average processor %	The average recent CPU percentage for this service instance, calculated over 30 seconds by default. Set the time period in `cpuAverageTimeSpan`, under `performanceCounterLogging`. See Spotfire.Dxp.Worker.Web.config for more information. (For information about tracking the average percentage of CPU usage for a service, see Monitoring CPU usage by services.)
Avg. disk queue length	The length of the queue for disk input-output. This number should be low.
Current processor %	The processor usage for the web client process. (For information about tracking the percentage of CPU usage for a service, see Monitoring CPU usage by services.)
Current time	The time (in UTC) when the page was last updated.
Data engine active queries	The number of active data engine queries. This value should not be far above 0 for very long. Normally, data engine queries do not take very long.
Data engine memory	The amount of memory used by the data engine. This includes all data views and data tables. If this value is a large part of Total working memory, then you might have performance problems related to RAM. If this value is only a small portion of the Total working memory, then you might have performance problems related to .NET memory. See Performance troubleshooting for more information.
Data engine paged in memory	The accumulated amount of paged in memory. This value must be much smaller than Data engine paged out memory
Data engine paged out memory	The accumulated amount of paged out memory. This value can be high, as long as Data engine paged in memory is much smaller.
Data engine queries finished	The number of finished low level data engine queries.
Data engine query cache memory	The amount of memory used by the data engine cache. This value can be very high without causing problems because it can be paged out to disk if necessary.
Dispose tracker finalizations
Failed queries	The number of failed data access queries.
Health status	The health status of the system: Available (or OK), Strained, or Exhausted
Idle threads in thread pool	The number of idle threads in thread pool (.NET) that are ready to be used.
Image render executions	The number of image render executions. Typically one image corresponds to one visualization.
Long running queries	The number of long running data access queries.
May be recycled	Depending on settings for `recoverMemory` and the current system status, the service instance may send an event to the server that it may recycle the service instance. For more information on `recoverMemory`, see its entry in Spotfire.Dxp.Worker.Web.config.
Memory health status	According to configured memory limits, this value is displayed as one of the following: `0:OK`. Indicates that the instance is under no pressure. `5:Strained`. Indicates that the instance is under pressure but is not a problem. `8:Exhausted`. Indicates that the instance is under a higher load, so avoid routing new users to this instance, but current users can keep working in this instance. Users of analyses in scheduled updates can be routed to a service instance with a status of `8: Exhausted`. If you discover that service instances that are used for scheduled updates are often in this state, you should consider adding more service instances to the resource pool. The health status is sent to the server to be used for routing decisions. For example, you want to avoid sending many users to service instances that are under a higher load. The limits that determine the health status are configurable for both CPU and memory.
Memory in all .NET heaps	The total MBytes in all .NET heaps, based on .NET CLR Memory.
Network kBytes/sec	The current rate of the network traffic, as measured in kilobytes per second.
Number of running service instances	The total number of running service instances on the server.
Number of sharable master documents	The number of loaded documents (analyses) that can be shared between users.
Number of users	The number of logged in users.
Process Private memory	The memory allocated for a process which cannot by shared by other processes.
Processor health status	The same as Memory health status above, but for the CPU load.
Thread pool queue length	The queue length for the thread pool (in .NET).
Total average processor %	The average recent CPU percentage for the node manager, calculated over 30 seconds by default. (For information about tracking the average percentage of CPU usage for a service, see Monitoring CPU usage by services.)
Total processor %	The total processor usage. (For information about tracking the percentage of CPU usage for a service, see Monitoring CPU usage by services.)
Total thread pool requests finished	The total number of thread pool jobs finished (.NET thread pool).
Total working memory	The amount of memory used by the web client process. If this value is high compared to Available memory, you might have performance problems related to RAM. See Performance troubleshooting for more information.
Uptime	The number of seconds since the service instance was started.