A customer is monitoring a lot of processes which are known to stay alive for a very short time sometimes. A good example is the iexplorer.exe, which sometimes starts only for 1 sec and then stops.
The aternity agent will try to inject to this process when it gets started, but before the agent is done with it, the process to inject into is already stopped.
The result: an error in the agent event logs, telling "couldn't inject into iexplorer.exe"
which is correct - but it is not an error which the customer should mind
That customer gets such errors very often, and I think there should be a way to acknowledge/ignore them, as they are irritating.