I had a sensor go bad and report a constant '100' for a couple of weeks. In the PI archive there is a event recorded ~20 seconds before the sensor went bad, followed by a single '100' value event, then nothing until the sensor was repaired. the next even retained in the archive had a value of 0.41.
When querying the value during the time the sensor was bad, I am seeing interpolation from 100 down to 0.41 over the whole span of weeks, when the sensor value was at 100 the whole time. My PI server seems to have missed recording events with the value still at '100' just before the sensor was repaired, missing the inflection point.
For stagnant tags I expected the point to be recorded as per:
For a stagnant tag in 'advise' mode, the interface will send the previous value it received and attach a current timestamp to it once the exception max time has been reached. For example, if the value and timestamp of an 'advise' tag remain constant on the OPC server and the tag has an ExcMax time of 1 minute, you will see new events with the same value but updated timestamps arrive at the PI Data Archive every minute. Depending on compression settings, these events may not be archived, but they will be received by the snapshot subsystem, which can be viewed in PI SMT > Data > Current Values. Keep in mind that this behavior is specific to the PI OPC DA Interface and is not relevant for other interfaces'
There was no disruption to the OPC connection, the point in question is read over OPC from a PLC with other points recorded the entire time. My understanding is that the Exception settings should have seen an event sent to the snapshot every 10 minutes with an updated timestamp. The Compression settings should have seen all of those dropped except for one event every 8 hrs. But it seems none of these Max Time events were sent to the archive.
Could anyone point me in the right direction to resolve this? I'm assuming here it is a configuration setting I've missed?