Yes - good questions. Packets out was un-affected. The only chart that shows a dip is Bytes/second out so sorry (Tx). But the Sar data is conclusive - it shows about the same throughput as Ganglia reports but does not show any big drops.
I was thinking maybe ganglia is getting one bad data point and it's throwing off the chart? But if Packets/Second is not effected then it makes this even stranger. BTW - All other aspects have been verified. System level and network level. Everything is coming back working optimally. One thing to note is we are using a bonded interface for fail over. I was thinking perhaps this might be screwing up how ganglia is calculating the values. But bonded interface shows up as bd0 and the primary as eth0 so there should be no confusion. The bdo is not active it's in hot failover mode.. Gmond - 3.0.1-1 Gmetad - 3.0.1-1 OS Distro - RHEL ES3 Linux 2.4.21-32 -- cp On Thu, 2005-11-03 at 03:22 -0800, Martin Knoblauch wrote: > Hi, > > hmm. "interesting". Was the Bytes-Out the only metric showing problems > at that time? What about Packets-Out. > > Loss of metric data is not unheard of, but only one metric affected is > strange. > > What platform and version (gmond, gmetad web-frontend) are you > running? > > Cheers > Martin > > --- "G. Francisco Perin" <[EMAIL PROTECTED]> wrote: > > > I am having a strange issue with ganglia reporting (not reporting) > > network traffic on a high volume web site. The graphs are reporting > > points of zero (0) data when SAR data is not showing the same > > information. Its disconcerting because if Ganglia is reporting good > > information then I have a problem. But all other indications are > > that > > things are fine. > > > > Here's an example of the SAR data for the interface: > > > > Time Rx Tx > > 05:14:00 845.78 945.85 > > 05:15:00 722.51 816.59 > > 05:16:00 752.33 840.65 > > 05:17:00 796.62 886.42 > > 05:18:00 888.71 990.34 > > 05:19:00 802.17 891.59 > > 05:20:00 760.13 851.22 > > 05:21:00 797.95 908.55 > > 05:22:00 909.58 1009.56 > > > > And attached is a graph from ganglia during the same time period. > > Notice there is a big drop in RX between 5:15-5:20? > > > > Any ideas what might be causing the chart to do this? Do you think I > > am > > looking at a real problem or something with ganglia reporting? > > > > -- > > cp > > > > > > > > > ------------------------------------------------------ > Martin Knoblauch > email: k n o b i AT knobisoft DOT de > www: http://www.knobisoft.de > > > ------------------------------------------------------- > SF.Net email is sponsored by: > Tame your development challenges with Apache's Geronimo App Server. Download > it for free - -and be entered to win a 42" plasma tv or your very own > Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php > _______________________________________________ > Ganglia-general mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/ganglia-general > >
