I have been using 2.2.3 running on CentOS 5.1 for a few months and have 
discovered a bunch of Cisco and Windows devices.  

Fine. 

Now, I still don't really understand what auto-discover really means in 
Zenoss-land.  I think what it means is that any device that Zenoss is told to 
discover (by adding the devices or selecting a subnet and clicking "Discover 
devices") is queried by Zendisc and and other appropriate processes which then 
model interfaces, CPU, disk systems, etc.  The zAutoDiscover property of a 
subnet doesn't really seem to do much.  All of my subnets have this property 
set to True and devices do not get discovered.

Here's how I think auto-discovery should work (basically, this is the HP 
Network Node Manager discovery model):

Any subnet that has zAutoDiscover set to True should be periodically scanned 
for live IPs and, at my option, be modeled by Zenoss.  Ideally, I would like 
Zenoss to only add devices that have a particular SNMP config (correct 
community names in SNMP v1, for example) and Zenoss should completely ignore 
all other devices.  I would also like Zenoss to be configured to limit what IP 
ranges can be discovered at all.  Ideally, devices would be automatically 
placed in appropriate classes based on the format of their device name.  The 
nice thing about this approach is that, if I control the SNMP configuration of 
devices, discovery and classification is completely hands-off.

Am I missing something?  Can Zenoss do what I want (Core, Enterprise, Service 
Provider)?

Matt




-------------------- m2f --------------------

Read this topic online here:
http://forums.zenoss.com/viewtopic.php?p=29190#29190

-------------------- m2f --------------------



_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users

Reply via email to