Remember, this all started from noticing that a network configuration that might even be common on Solaris can cause a mixture of short and long hostnames when gmond does the name lookup from the IP in the multicast data. If your primary query node goes down or is unreachable for some other reason then gmetad will start seeing new hosts suddenly appear because the secondary gmond will get a few different hostnames than the primary got. Again, I don't really care about the exact lookup method as long as ganglia makes sure to get consistent hostnames.
~Jason On Wed, 2003-03-26 at 04:55, Leif Nixon wrote: > "Jason A. Smith" <[EMAIL PROTECTED]> writes: > > > As long as the hostnames that ganglia gets from the IPs are consistently > > in the fqdn format then it would be better than the current situation > > which results in a mixture of short and full names. > > That isn't possible, in the general case. FQDN is a DNS concept, and often > cluster nodes doesn't have anything to do with DNS; they don't use DNS > for name resolution, and their IP address don't appear in any DNS zones. > > > Maybe a check to verify that the hostname lookup returned a fqdn, if > > it didn't then add the domainname to it. > > Similar problem here; cluster nodes often don't have any concept of > a "domain". -- /------------------------------------------------------------------\ | Jason A. Smith Email: [EMAIL PROTECTED] | | Atlas Computing Facility, Bldg. 510M Phone: (631)344-4226 | | Brookhaven National Lab, P.O. Box 5000 Fax: (631)344-7616 | | Upton, NY 11973-5000 | \------------------------------------------------------------------/
