We had a strange experiende. One of our sites send a burst of future data into the system. The buffermanager showed an error and the points which had future data. At the same time the number of events pr second in the pibuffermanager dropped almost to a third of the normal value?
We did a "net stop pisnapss" and then a pisnapss -f and waited for the clearance message in the log and then a "net start pisnapss". After this everything went back to normal
My question: Why was the events pr second much lower when we had the error? Did we loose data. I presume that compression is off while the error is there. Is that true?