"absolute values" in RRD graph

[ipnd-alerts]

Hi,

I want to get statistics about up/down time on our Solaris cluster in a 
graph form.
I would like to see 4 states in a graph.
3 - UP
2 - FAILED OVER
1 - PLANNED DOWNTIME
0 - DOWN

I got the cluster script that reports these states to BB or Xymon.
But I can't find a way to do the graph.
RRD graphs seems to always average out data.
The problem is that I want to report the absolute values and not the 
averaged out data that seems to be default.
For instance if the cluster is DOWN ( 0 ) for 10 minutes I would like 
that state to remain and not to averaged out to 2.5 after an hour.

Is this possible?

- Roland