Hi again, I just pulled the latest snapshot from master, did some tests and here are the results:
submitted counter when ES works well (both batch and no batch) - works ES is down => failed.http and failed.httprequests remained 1. The number of "submitted" remained the same. This was valid both when I used an action queue (with action.resumeretrycount="-1") and when I didn't. ES throwing mapping errors => the submitted counter kept increasing, the failed.* counters remained 0. Again, this happened with both bulkmode enabled and disabled. I'll happily do some more tests (maybe do a dirty script after the next round), but if you need any help with the test env (like a step-by-step instruction), let me know. I can even provide you a remote ES to play with (except you won't be able to shut it down :D), because we're getting ready to release the first beta of our logging SaaS called Logsene<http://sematext.com/logsene/index.html>, and it exposes the ES API. Guess what we use for syslog reception? :D Best regards, Radu 2013/10/17 Radu Gheorghe <[email protected]> > Thanks, Rainer! I wrote it down in my todo list. Will have a look soon and > come back with feedback. > > Best regards, > Radu > > > > 2013/10/16 Rainer Gerhards <[email protected]> > >> On Wed, Oct 16, 2013 at 5:51 PM, Rainer Gerhards >> <[email protected]>wrote: >> >> > >> > On Wed, Sep 18, 2013 at 7:45 PM, Radu Gheorghe <[email protected] >> >wrote:- >> > failed.es counter doesn't seem to work. I mean, at least not for JSON >> > >> >> errors, which was the easiest way to poke it. I also tried a slightly >> >> smarter hack of making ES think the message field is numeric, to cry a >> >> NumberFormatException when it finds out it's a string. Sadly, >> >> failed.esremained 0. Anyway, here's how I did that: >> >> >> > >> > I still need to look into that but thought it is better to least fix the >> > other one ;) >> > >> > I'll keep you posted. If you would llike to give the new version a try, >> > that would be great. >> > >> >> OK, I couldn't stand it and had another look. It seems that only failures >> in batch mode were counted. If so, I have fixed this now. Please note that >> this counter does not count messages but actully failure occurences. I >> thought about message couting, but that's more performance intense, so >> probably not something we want to do (at least by default). This is >> documented in the stats counter description now. >> >> I did *not* these fixes other than by compiling -- due to lack of readily >> available environment. >> >> So feedback would be most appreciated. >> >> Rainer >> _______________________________________________ >> rsyslog mailing list >> http://lists.adiscon.net/mailman/listinfo/rsyslog >> http://www.rsyslog.com/professional-services/ >> What's up with rsyslog? Follow https://twitter.com/rgerhards >> NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad >> of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you >> DON'T LIKE THAT. >> > > _______________________________________________ rsyslog mailing list http://lists.adiscon.net/mailman/listinfo/rsyslog http://www.rsyslog.com/professional-services/ What's up with rsyslog? Follow https://twitter.com/rgerhards NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE THAT.
