There needs to be a 64-bit version of the CatGH5 interface that is capable of parallel processing in order to efficiently process an quantity of data/files (that the 32-bit version is not up to the task of handling well).

Currently only POINT/ANALOG fields support COUNT. This request would be to enhance the interface to additionally support the COUNT field.

PI-SNMPTrap interface currently only has support for SNMPv1 and SNMPv2C trap. Feature request for PI-SNMPTrap interface to include support for SNMPv3 trap.

Monitoring of snmp traps originating from CISCO switch devices require support for SNMPv3 trap.

We would like the PI CAISO interface to parse through the .XML files and exclude XMLs that contain certain specified <Batch Types>. Only certain batch types contain valid data for consumption.

Add support for Conversion Factor PI Point attribute to PI Interface for BACNet.

Note: This was previously Enhancement 118928.

If the RTU provides a timestamp of 0 when using RTU provided timestamps, which does not conform to the DNP3 spec, allow the option to timestamp the event with PI time or drop the event altogether.

Note: This was previously Enhancement 123208.

When using the interface to gather data for hundreds of SNMP devices, managing the .pwd file becomes very burdensome due to an inability to format the file. Comments are helpful in sorting devices, but they are removed upon encryption. It would be nice to see a tool that helps with the management of passwords.

When validating/troubleshooting data flow issues, it would be helpful to see metrics on which interfaces are actively sending data to the buffering subsystem via Buffering Manager. If an interface is in a bad state or stopped, it would be obvious which interface is having issues since the count would be zero. Currently the counts in the buffer queue are for all interfaces, so this only gives the user an idea on whether the buffer subsystem is forwarding data to the data archive, without knowing which interfaces this data is coming from.

PI Buffer Subsystem must have a option to set up the user account is logged instead of using Windows Domain or Local account that is running the windows service. This is a huge security issue when using Buffer on client computer to send manual data because you will need to create a Mapping or Trust and give permission in so many pi points when the user logged on computer should have access to write in few pi points.

Please prioritize this one!

Add functionality for the Davis to Pi Interface to read additional sensor data points. The Davis to Pi Interface does not support the use of additional sensors, such as a second or third temperature sensor. The Davis console supports additional sensors, but the Pi Interface cannot read them.

Failover mechanism in fxbais (Schneider-Electric PI interface for Windows) doesn't lead to a failover when the primary interface node loses connection to the network (f.i. when its NIC fails). It just starts buffering and sends its data after the connection is back again.
It keeps on updating its watchdog CALC block on IA, so no take over is initiated.
Some applications require "near real time" data, so if it comes in a day later, it's not very useful.
In such a case, the secondary interface node could perfectly take over the data collection.

Scenario: I have a tag that pings site A. I have the response times, which quantifies the response times of the network, but this line is also a bit slow and we suspect packet loss, but I want that in a tag.

Enhancement for Ping Interface to add -f switch 'Set Don't Fragment flag in packet. When sending large packets we do not want to Fragment.

Case: When monitoring a site ping worked for small packets, however with large packets sent, the site showed packet loss which in turn was an actual hub issue missed when using Ping Interface.

Right now PerfMon only supports ArcNoReplace. PerfMon has a history of failing over if a server is unreachable. And since PerfMon only supports Hot failover, every time it fails over the new interface tries to write values that the old one already wrote. I would like to be able to specify ArcNoReplace so I don't get an error flood every time the interface fails over.

Currently DNP3 outputs are done one event at a time. We could improve performance by either supporting multiple DNP messages per TCP packet or multiple analog output commands per DNP message.

The newest version of UCOS now supports string tags.  Currently the CSIUCOS interface (1.0.0.22 as of writing) does not support these types of tags.