LMS 3.0 Device Discovery only works after manual addition

Unanswered Question
Jan 18th, 2008

Hi, I am trying to discover my devices, but having mixed luck. I discovered one batch of devices fine after setting the seed device to their upstream routers and setting the IP range. Then I tried a second batch and got zero devices discovered.

I logged into the first device on the list, term mon'd it and ran device discovery again. Nothing, no log messages, no devices discovered.

Then I added the first machine manually and named it "test", then ran discovery again. This time it found it fine, just couldn't log in (default credentials routinely fail until I update them for new machines with the same information, but that's another, more manageble, problem). I re-applied the default password in CM and the box was discovered without any errors or problems.

My main problem is obviously that I don't want to manually add 600 devices just for auto-discovery to find them. But the more immediate problem is that I don't see any error messages in LMS that can clue me in to what the problem is. I've looked through this forum, but I don't see anything helpful in the CSCOpx\objects\smarts\local\logs files. Can someone point to error logs within the GUI or on the system? I've been poking around docs and the filesystem to no avail.

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Joe Clarke Fri, 01/18/2008 - 13:13

Once a device is discovered (and it is reachable), it is added to DCR. It doesn't go away after that unless it is manually deleted. Once the device is added to DCR, then other applications like RME, DFM, and Campus can manage that device.

If Discovery is not finding this second batch of devices, then there must be an SNMP problem as Discovery only uses SNMP to communicate with the devices. To debug this, go to Campus Manager > Admin > Debugging Options > Device Discovery, and enable debugging for the devdiscovery module. Run a new discovery, and post the discovery.log.

ktokashhh Fri, 01/18/2008 - 13:40

Thanks for the quick response. I enabled debugging and re-ran the failed scan, then re-ran the successful scan. I also watched logs just in case some acl got changed or something, but no action there. I don't see any difference between the log output below, and the log4j system error is the only thing that looks suspicious. I'm not sure what that would result in though.

**************************************

FAILED DEVICE DISCOVERY - 0/10 DEVICES

**************************************

2008/01/18 13:15:33 main MESSAGE DiscoveryMain: Properties will be read from C:\PROGRA~2\CSCOpx\campus\etc\cwsi\DiscoveryStatus.properties

2008/01/18 13:15:33 main MESSAGE DiscoveryMain: Properties will be read from C:\PROGRA~2\CSCOpx\campus\etc\cwsi\DiscoveryStatus.properties

2008/01/18 13:15:33 main MESSAGE DBConnection: Created new Database connection [hashCode = 7756310]

inserting record ... INSERT INTO SystemStatus (processid, starttime, status, count, othercount) VALUES ('0', '2008-01-18 13:15:34.735', '1', '0', '0')

getting dbid query ... SELECT dbid FROM SystemStatus WHERE processid='0' AND starttime='2008-01-18 13:15:34.735'

Common trust user is: admin

User name in Security context is: admin

log4j:ERROR No appenders could be found for category (CTM.common).

log4j:ERROR Please initialize the log4j system properly.

2008/01/18 13:15:35 main MESSAGE DCRDevWrapper: Fetching DCRDeviceIDs

2008/01/18 13:15:41 main MESSAGE DCRDevWrapper: Finished fetching DCRDeviceIDs

2008/01/18 13:15:41 main MESSAGE DCRDevWrapper: Fetching DCRDevices

2008/01/18 13:15:41 main MESSAGE DCRDevWrapper: Finished fetching DCRDevices

2008/01/18 13:15:41 main MESSAGE DiscoveryMain: Properties will be read from C:\PROGRA~2\CSCOpx\campus\etc\cwsi\DiscoveryStatus.properties

2008/01/18 13:15:41 main MESSAGE DiscoveryMain: Properties will be read from C:\PROGRA~2\CSCOpx\campus\etc\cwsi\DiscoveryStatus.properties

2008/01/18 13:15:41 main MESSAGE DCRDevWrapper: Closing DCRProxy

Calling update in discovery

updating status =

dbid = 929

processname = Discovery

startTime = 2008-01-18 13:15:34.735

endTime = 2008-01-18 13:15:41.769

otherCount = 0

status = Idle

count = 0

updating record ... UPDATE SystemStatus SET endtime='2008-01-18 13:15:41.769', status='0', count='0', othercount='0' WHERE dbid='929'

ktokashhh Fri, 01/18/2008 - 13:40

**************************************

SUCCESSFUL DEVICE DISCOVERY - 13/13 DEVICES

**************************************

2008/01/18 13:29:29 main MESSAGE DiscoveryMain: Properties will be read from C:\PROGRA~2\CSCOpx\campus\etc\cwsi\DiscoveryStatus.properties

2008/01/18 13:29:29 main MESSAGE DiscoveryMain: Properties will be read from C:\PROGRA~2\CSCOpx\campus\etc\cwsi\DiscoveryStatus.properties

2008/01/18 13:29:30 main MESSAGE DBConnection: Created new Database connection [hashCode = 7756310]

inserting record ... INSERT INTO SystemStatus (processid, starttime, status, count, othercount) VALUES ('0', '2008-01-18 13:29:31.332', '1', '0', '0')

getting dbid query ... SELECT dbid FROM SystemStatus WHERE processid='0' AND starttime='2008-01-18 13:29:31.332'

Common trust user is: admin

User name in Security context is: admin

log4j:ERROR No appenders could be found for category (CTM.common).

log4j:ERROR Please initialize the log4j system properly.

2008/01/18 13:29:32 main MESSAGE DCRDevWrapper: Fetching DCRDeviceIDs

2008/01/18 13:29:38 main MESSAGE DCRDevWrapper: Finished fetching DCRDeviceIDs

2008/01/18 13:29:38 main MESSAGE DCRDevWrapper: Fetching DCRDevices

2008/01/18 13:29:38 main MESSAGE DCRDevWrapper: Finished fetching DCRDevices

2008/01/18 13:29:44 main MESSAGE DiscoveryMain: Properties will be read from C:\PROGRA~2\CSCOpx\campus\etc\cwsi\DiscoveryStatus.properties

2008/01/18 13:29:44 main MESSAGE DiscoveryMain: Properties will be read from C:\PROGRA~2\CSCOpx\campus\etc\cwsi\DiscoveryStatus.properties

2008/01/18 13:29:44 main MESSAGE DCRDevWrapper: Closing DCRProxy

Calling update in discovery

updating status =

dbid = 932

processname = Discovery

startTime = 2008-01-18 13:29:31.332

endTime = 2008-01-18 13:29:44.524

otherCount = 0

status = Idle

count = 0

updating record ... UPDATE SystemStatus SET endtime='2008-01-18 13:29:44.524', status='0', count='0', othercount='0' WHERE dbid='932'

Joe Clarke Fri, 01/18/2008 - 14:17

I do not see in either of these logs that devdiscovery debugging is enabled. If it is, then there are no devices in DCR, and no seed devices configure for discovery.

ktokashhh Fri, 01/18/2008 - 14:16

Ok, mystery solved. I misunderstood the discovery process. We have a layer-2 distribution layer and I didn't include it in the ip range to discover. I assumed that the discovery process would recurse anyway, which it does not. Fair enough, either way is fine with me.

Thanks for the log tips, I was tinkering with the different debug outputs and went down this path.

*edit*

I should mention that the reason the discovery succeeded in the other case was that I tried it in a small POP, where we have a collapsed core/distribution. All the edge devices I wanted to discover are one L2 hop away from the seed devices. :)

Actions

This Discussion