On Wed, Jun 25, 2008 at 1:40 PM, Kirk McDonald <[EMAIL PROTECTED]> wrote: > On Wed, Jun 25, 2008 at 12:26 PM, Bernard Li <[EMAIL PROTECTED]> wrote: >> Hi Kirk: >> >> On Wed, Jun 25, 2008 at 12:01 PM, Kirk McDonald >> <[EMAIL PROTECTED]> wrote: >> >>> I am running Ganglia 3.0.7. Let's say I have a large number of hosts. >> >> Since you say you have a large number of hosts, perhaps you are >> running into I/O issues with rrd. >> >> I'd suggest running iostat to see what is going on, and if there is a >> lot of disk activities consider putting /var/lib/ganglia/rrds on >> tmpfs. You'll find a lot of information if you google for this. Lots >> of hidden gems in the mailing-list archives available in >> mail-archive.com as well. >> >> If that's not the case, well, it must be something else then. >> >> Cheers, >> >> Bernard >> > > As it happens, I tried this already. There was no substantial change. > > -Kirk >
I take this back. I am not sure what I thought I saw with my earlier test, but looking at the configuration using tmpfs again, it does not suffer from hosts being marked as down. gmetad's TN values still average higher than those given by the gmonds, but the difference is much smaller. It is definitely a disk I/O problem. -Kirk ------------------------------------------------------------------------- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php _______________________________________________ Ganglia-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ganglia-general
