Re: [Ganglia-general] Fine grained monitoring problem

2016-04-11 Thread Karol Korytkowski
urceforge.net > From: vli...@veus.hr > Date: Fri, 8 Apr 2016 09:47:40 -0400 > Subject: Re: [Ganglia-general] Fine grained monitoring problem > > > Hi Theofilos, > > by default gmetad polls gmonds every 15 seconds and inserts them into > RRDs. You can attempt to change tha

Re: [Ganglia-general] Fine grained monitoring problem

2016-04-11 Thread theofilos moyratidis
@lists.sourceforge.net From: vli...@veus.hr Date: Fri, 8 Apr 2016 09:47:40 -0400 Subject: Re: [Ganglia-general] Fine grained monitoring problem Hi Theofilos, by default gmetad polls gmonds every 15 seconds and inserts them into RRDs. You can attempt to change that by adding

Re: [Ganglia-general] Fine grained monitoring problem

2016-04-08 Thread Vladimir Vuksan
Hi Theofilos, by default gmetad polls gmonds every 15 seconds and inserts them into RRDs. You can attempt to change that by adding the polling interval e.g. 1-3 seconds in gmetad.conf. To do so you would add it after the grid name e.g.

[Ganglia-general] Fine grained monitoring problem

2016-04-08 Thread theofilos moyratidis
I have a cluster and I want to get metrics every 1-3 seconds, to benchmark some algorithms on spark. I’ve set the data_source “my cluster name” 1 hosts…. (multicast). The UI on the hours tab, gives me the results in every 12 seconds, that’s too chunky for what I want to accomplish. I used the