Hi Joel, The mbean approach depends upon storing traffic stats over a period of time, so is doable but complex. I am thinking about another approach to check logfiles time-stamp (to check producer activity) in conjunction with the mtime for topic offsets (consumers activity). So if both timestamp and mtime are older than 'n' days that may imply a defunct topic. Will appreciate your feedback.
Thanks, Jagbir On Thu, Feb 5, 2015 at 6:57 PM, Joel Koshy <jjkosh...@gmail.com> wrote: > There are mbeans > (http://kafka.apache.org/documentation.html#monitoring) that you can > poke for incoming message rate - if you look at those over a period of > time you can figure out which of those are likely to be defunct and > then delete those topics. > > On Thu, Feb 05, 2015 at 02:38:27PM -0800, Jagbir Hooda wrote: >> First I would like to take this opportunity to thank this group for >> releasing 0.8.2.0. It's a major milestone with a rich set of features. >> Kudos to all the contributors! We are still running 0.8.1.2 and are >> planning to upgrade to 0.8.2.0. While planning this upgrade we >> discovered many topics that are no longer active and best be deleted. >> Since it would be a common task faced by many kafka adopters I thought >> I'd raise it here and seek expert advise. Basically what we desire is >> a utility/script/steps to first identify the defunct topics (let's say >> those topics that haven't seen any traffic in the past 'n' days) and >> then delete them. Will appreciate your response. >> >> Thanks, >> Jagbir >