LMS 4.1 W2k8 R2
I recently installed 17 stacks of 3560V2 switches at a site and added them to LMS4.1. What I have found is that data collection is failing for all but the top switch in the stack. In other workds, in some cabinets I may have 4 switches connected back to a 6500 core. Data collection succeeds for only the switch connected to the core. UT info is not collected for the switches that fail data collection.
If I go to the device status dashboard I see that I have data collection succeeding for 56 devices and failing for 96 devices. For the devices that fail some are listed as unreachable and others as never-reachable. In either case, the following is suggested as a resolution:
Verify the following :
1. Device is reachable using ping.
2. Verify that DCR is configured with correct credentials for the device.
These devices are reachable. Inventory collection is succeeding. If I test reachability in device center I get successes across the board - even for SSH, though no username is specified in DCR so I can't see how that works.
The 17 stacks I just installed are configured as clusters in those cabinets that have more than 1 switch. However, they are not managed as clusters by LMS. They all have IP addresses. I just use clustering to facilitate bouncing around a stack from a telnet command line.
I deleted one of these switches and re-added. Now ping, HTTP, SSHv1 & v2 succeed in device center - but everything else fails. When I run a credential verification job on this switch it reports SNMP read and write success, telnet & enable success, and SSH "no value to test". The credential verification results are as I would expect.
I am at a loss at this point. How do I troubleshoot the data collection issue?
-Jeff
