Hi,
we have a temperature probe that in normal conditions measures temperatures in teh range of 15 - 30°C.
Due to a bad connection the temp sensor is showing readings of 850°.
The issue is repaired now, but consulting trend data now results in 'unreadable' graphs as the system scales to show the peaks.
In the era of mysql it was possible to remove those incorrect data entries from the trend database.
Is this possible on a cassandra system as well??
Hi Hans,
It's probably possible on Cassandra to remove this, but as that is not my domain of expertise, I cannot advise on that (there are trend window calculations done min/max/avg, don't know how tampering with the data will influence that or how trend points relate to each other).
Anyhow, about the 'unreadable' graphs: it is possible in Cube to right-click on the Y axis and select 'Y-axis settings...'. Instead of setting it to "Auto", a "Range" can also be specified where the maximum is set to e.g. 50 and the minimum to 0. This way the 850 peak will not be displayed.
Hope this helps a bit to have more readable graphs again.
Regards,
Agree on the trick to clean up the display – but still that keeps the excessive spikes in the db for avg/min/max calculation.
I guess recreating the element will be the easiest way out (at the expense of losing the history).
I’m not sure why you have to recreate the element. In any case when you zoom or pan past the point where the unexpected value happened, the auto scaling will adjust and you’ll have a nice graph again.
Scaling is there, however the Average and Max values are messed up.
I see that this question has been inactive for some time. Do you still need help with this? If not, could you select the answer that has been most helpful for you (using the ✓ icon)?