When forwarder's queue is maxed out, data [ monitor files or directories ] will get backed up. What is "backed up" ?

[SamChang]

Hi Guys ...

I get a question about input queue on forwarder ?

I found a document "Use persistent queues to help prevent data loss"

I see below description in that document.

By default, forwarders and indexers have an in-memory input queue of 500KB. If the input stream is running at a faster rate than the forwarder or indexer can process, to a point where the queue is maxed out, undesired consequences will occur. In the case of UDP, data will drop off the queue and get lost. For other input types, the application generating the data will get backed up.

I don't understant what is "backed up" ?

If my forwarder gets data from directories ...

"backed up", It mean ...

forwarder continually reads log form files and write log to local hard-disk, RIGHT ???

Please help me to get correct anwsers or documents, thanks

[lukejadamec]

By 'backed up' they mean the log for the application will fill, but Splunk will not read it. Splunk will remember where it left off, but when it starts reading the log file again there will be a backlog that needs to be read.

So long as the log is not deleted, over written, or rolled (if you're not monitoring rolled files), then you should not have a problem.