Hi,
I don't see any similar problem on machines upgraded to Centos5.3 that are monitored with (and running) OpenNMS, so I'd guess that since you didn't change your snmpd.conf settings it is MRTG-specific.
I think it's my server, quite possibly I screwed up something during the initial setup two years ago or along the way updating it from 5.0 and so forth until it's not behaving in any recognizable manner anymore.
And btw: OpenNMS might be overkill for your purpose, but you might want to take a look: http://www.opennms.org.
It looks good and I decided to give it a try in hope that maybe it can be up and running faster than I can get MRTG to work again. Unfortunately, as above mentioned, my server does not behave like a CentOS server anymore. Following the steps at OpenNMS, I get to the install -dis stage where it promptly dies because it cannot find jrrd.
downloaded jrrd but it refuses to ./configure because it cannot find rrd_create
yum install rrdtool but there was no rrd_create
searched online and the only result that was similar... was somebody having the same problem on a Solaris server <-- hence making me wonder if I was logging into the wrong server. Using the instructions there however, I at least learnt how to tell configure where rrdtool was... but it still cannot find rrd_create for the ./configure process
Having spent almost 5 days on this, I'm officially giving up on monitoring the server with these tools. Writing a PHP script seems a lot faster, I've already gotten a basic script running to pull load figures from exec'ing uptime and emailing warnings if the load figures stay above a certain level.
Now I just have to expand the script to exec snmpget for the other metrices I need to keep track of. It's really frustrating that I have to resort to writing my own code when these things worked fine for other people.