You could, but I think it would be a major rewrite of how the data is pulled and processed... You'd probably have to create wrapper scripts around Nagios Plugins or your own SNMP Walk command based data sources, and then do the logic for what return value is too high...
This is because as I understand it, the thresholds which give you the high utilization are part of the graph definitions... -- James Pulver Information Technology Area Supervisor LEPP Computer Group Cornell University pearsch1 wrote, On 9/2/2009 3:23 PM: > We use Zenoss for monitoring ~250 devices. These servers are a mix of > Windows, Linux, and network devices. The network devices all use SNMP, and > the servers use both either WMI/SNMP or SSH/SNMP. > > The server is a VM with two cores and about 1.5GB of memory. > > The server occasionally gets very sluggish, so I'm trying to tweak its > performance. The disk appears to be the bottleneck. When the web interface > gets very sluggish, the total IO from iostat goes to 100%, and context > switches from vmstat sometimes go over 2000. > > In order to increase the disk performance, I would like to disable the > collection of much of the rrd data that is collected on our servers (the data > that appears under the perf tab), as we don't use this data. We would, > however, still like to receive notifications of high disk utilization, high > cpu utilization, etc. I just don't want the historical performance data > written to disk. > > Is there a way to do this? > > > > > -------------------- m2f -------------------- > > Read this topic online here: > http://forums.zenoss.com/viewtopic.php?p=38797#38797 > > -------------------- m2f -------------------- > > > > _______________________________________________ > zenoss-users mailing list > zenoss-users@zenoss.org > http://lists.zenoss.org/mailman/listinfo/zenoss-users _______________________________________________ zenoss-users mailing list zenoss-users@zenoss.org http://lists.zenoss.org/mailman/listinfo/zenoss-users
