I run TailFiles input for a dir where I have about 3k files generated by MS Exchange.
TimberWinR went into high cpu and mem usage and what I concluded it was trying to read all the files from beggining.
This is ok for regular file input method but should not be working like that if we call this "tail".
IMHO the behaviour should be to seek to end of file and read only new lines.
I run TailFiles input for a dir where I have about 3k files generated by MS Exchange.
TimberWinR went into high cpu and mem usage and what I concluded it was trying to read all the files from beggining.
This is ok for regular file input method but should not be working like that if we call this "tail".
IMHO the behaviour should be to seek to end of file and read only new lines.