I am trying to deploy the anomalydetection command and get the following warning:
'anomalydetection' command: limit for values of field '_raw' reached. Some values may have been truncated or ignored.
Now, the base search itself returns around 3500 events back, so I do not think this is the problem (number of events).
Some of the events are quite big though as they relate to the DB errors and sometimes there are the SQL strings in them.
So, if this is not about the number of events but just that some of them get truncated when being processed by the anomalydetection, then I am fine with that.
Could you please confirm that I interpret the above warning correctly?
Is there is any limitation in respect to the number of events as an input for anomalydetection?
If yes, how would I increase this?
At the moment I am evaluating the errors only 7 days back, but the plan is to search for the last 180 days, so I guess there will be hundreds of thousands of the events returned.