cactid fails when pinging certain hosts

[coreone]

I am having a two-fold issue with a new cacti installation involving gaps in the graphs. My current setup is as follows:

Dell PowerEdge 1850, 2G RAM, dual 3.0GHz processors
Mandriva 10.2
Apache 2.0.54, PHP 5.0.5, net-snmp 5.2.1.2
cacti 0.8.6g
cactid 0.8.6f-1

Now, my problem, firstly, is that I have been noticing that since I moved to this new server, I have been getting about 1/2 hour gaps in my graphs at sporadic times. I therefore decided to try cactid to see if that was an easy drop-in solution.

My next problem is that cactid seems to run much better for most instances, except that I now have the unfortunate side effect that 3 of my hosts always register as DOWN if I use cactid. All my other hosts work fine, but these 3 will not register as UP except if I change back to cmd.php. The hosts are (1) Netscreen 5XP firewall and (2) Tripp Lite SNMP/Web SOLO boxes. I am currently running Ping ONLY for Downed Host Detection and that seems to be where the host status is failing. Therefore, I'm getting gaps in my graphs due to the fact that cacti thinks they are offline.

So, now I am stuck because I can't figure out how to either fix the original gaps problem or how to get cactid to somehow realize that those 3 hosts aren't down all the time. I did an 'rrdtool fetch' on the .rra data sources and find that there are gaps in the data file itself, so obviously the data wasn't recorded during the poll.

Any help would be greatly appreciated. Thanks.