Please ensure Javascript is enabled for purposes of website accessibility
Powered by Zoomin Software. For more details please contactZoomin

Buffering and High Availability

PI Buffered Physical Servers statistics

  • Last UpdatedJul 15, 2025
  • 4 minute read

PI Buffered Physical Servers counters record the buffering status of physical computers. Counters starting with Autotune> or Post> are primarily for troubleshooting when working with Customer Portal.

Attribute

Description

Autotune> Countdown to restart

Autotune> Countdown to optimization restart (sec) - How long (in seconds) autotune will wait before it restarts autotuning if the number of events pending isn't below the value of Autotune events trigger (default value = 12,000,000). You can find Autotune events trigger in Buffering Manager's Buffering Settings window.

Note: This countdown timer will get reset each time:

  • The number of queued events goes below the previously lowest value

  • The timer expires and the number of events pending is less than or equal to the value of Autotune events trigger

Autotune> Events Pending Low

Autotune> Lowest number of pending events since last optimization. If the number of pending events gets below this, PI Buffer Subsystem is catching up.

Autotune> Events/send upper limit

Autotune> The upper limit of the number of events per send used by the autotuning algorithm. This may be reduced from the default value when there are errors (timeout, message too big, etc.) sending data.

Autotune> Mode

Autotune optimization mode (0 = off; 1=start; 2=init; 3=run; 4=optimized)

Autotune> Optimization Step

Autotune> Each time the autotuning algorithm tries a new value, this counter is incremented by one.

Autotune> Optimized events/send

Autotune> optimized number of events to send each time

Autotune> Queue File Count Low

Autotune> Lowest number of queue files since last optimization. If the number of queue files gets below this, PI Buffer Subsystem is catching up.

Autotune> Send Err Recovery

Autotune> Send error recovery mode (0 = false, 1 = true). If PI Buffer Subsystem is in send error recovery mode, then it will reduce the number of values it is sending to try and clear the error.

Autotune> Trial events/sec

Autotune> Events/Sec rate for the previous trial value of Events/Send

Autotune> Trial events/send

Trial value for the number of events to send each time

Health

Health of this server (0 = OK, 1 = Warning, 2 = Error, 3=Critical)

Last Successful Send Time

The last time successfully sent data to the server

Points With Post Error

Number of points received errors when posted to the server.

Post> Current events/send

Post> number of events to send each time

Post> Seek-Send-Peek Time

Post> The total time in milliseconds to seek past the events successfully sent to the server, asynchronously send events previously peeked, and seek the next events to send.

Note: Seeking and Peeking happen in parallel with sending the previously peeked events. So this counter's value is not the sum of the 'Post> Time to Seek', 'Post> Time to Send', and 'Post> Time to Peek' counter values.

Post> Send mode

Post> Send mode (0 = Not sending; 1 = Lock error; 2 = Created; 3 = Initialized; 4 = No events; 5 = Peeked first events; 6 = Finished sending and events peeked; 7 = Sending; 8 = Send error; 9 = Wait error; 10 = Send OK; 11 = Seeking sent events; 12 = Seek error; 13 = Seek OK; 14 = Peek error; 15 = Peek OK; 16 = Peeking next events; 17 = Sent events and peeked events before; 18 = Sent events and no more to send; 19 = Post OK; 20 = No events to seek)

Post> Time to Peek (ms)

Post> The time in milliseconds to peek events out of the queue including copying for peeking ahead

Post> Time to Seek (ms)

Post> The time in millisecond to seek past the events successfully sent to the server

Post> Time to Send (ms)

Post> The time in milliseconds to send the data to the server and get an acknowledgement

Queue> Capacity

The estimated time (in seconds) that PI Buffer Subsystem can buffer data for this server if connection to the server is lost.

Queue> File Count

Number of queue files

Queue> Reset Folder Count

Number of buffer queue reset folders currently in storage. If PI Buffer Subsystem detects the queue is corrupted and is not able to recover it, a new queue is created so that data can flow. Each time that happens, the old corrupted queue files are saved in a separate folder under the configured queue path.

Queue> Write Errors

Number of events failed to be written to the buffer queue.

State> Is Connected

Whether PI Buffer Subsystem is connected to this server (0 = No, 1 = Yes).

State> Is Queuing Enabled

Whether data queuing for this server is enabled (0 = No, 1 = Yes). Users can disable queuing by issuing the command pibufss -bc dismount. When queuing is disabled, data is not queued or sent to the server.

State> Is Sending Enabled

Whether sending data to the server is enabled (0 = No, 1 = Yes). Users can disable sending by issuing pibufss -bc stop. When sending is disabled, data is queued but not sent to the server.

State> Is Sending OK

Whether PI Buffer Subsystem is sending or ready to send data to this server (0 = No, 1 = Yes).

Total Events Sent/sec

Rate of events sent to this server.

Total Queue Writes Exclusive/sec

Rate of events written to this buffer queue but not other queues in the same collective (no fanning mode).

Total Queue Writes/sec

Rate of events written to the buffer queue for this server.

Total Queued Events

Total number of buffered events for this server.

Total Snapshot Posts/sec

Rate of data post calls to this server.

TitleResults for “How to create a CRG?”Also Available in