On Mon 2010.01.04 at 21:17 +0000, Stuart Henderson wrote: > On 2010/01/04 14:04, Okan Demirmen wrote: > > any reason to not have mrtg depend on p5-RRD? most people use mrtg with > > rrd's nowadays as opposed to rateup. > > not sure I agree 100% with "most people use mrtg with rrd's", > since it either involves running code from the webserver, or > writing your own script to generate the graphs which are done > automatically with the standard rateup-based version.. but
i haven't seen mrtg store data in rateup-databases in a long time, only in rrd databases; i wasn't referring to the rendering of graphs, just storage of the name/value pairs after polling. maybe there are folks stilling using rateup's format ;) but i see a point in that pulling in p5-RRD, and subsequent rrdtool dependencies, can imply that one should generate the graphs in the same place. > I think depending on p5-RRD probably makes sense. :) > > also remove the silly 7 second sleep about ordering a cd...the joke > > doesn't annoy me, but the sleeps do. > > emphatically YES! :-) heh... > while on the subject; other than the people I've already talked to > about it, is anyone seeing MRTG fail to fetch correct values via > SNMP since updating packages in the last month or two? (fetches > 0's, though it does show the router uptime correctly). i just started running mrtg on a new (-current) install and i've yet to see 0's, but i will look out for them; is it consistently zero for you? cheers, okan
