cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1467
Views
0
Helpful
6
Replies

Inventory not updated

raindrop18
Level 1
Level 1

I have RME 4.3.1 and some devices never collected/ no longer updated. the error I got from inventory collection job "Transport session to device failed. Cause: Authentication failed on device." but I didn't see any Authenication failure including run credential verification confirm that. also able to snmpwalk, ssh from CW. GUI and server. this device are mostly 2960. but not all I have hundreds of 2960 the same model/IOS version  they are updating inventory just fine but some not where/what else I can troubleshoot the problem. Thanks in Advance. P      "

1 Accepted Solution

Accepted Solutions

A wrong digest message means that the polling manager is using the wrong SNMPv3 password, or the wrong digest algorithm.  Since you said you can SNMP Walk the device using SNMPv3 from Device Center, I suspected the DCR credentials are wrong.  Therefore, I suggest you re-enter them in DCR (and verify the authentication algorithm).  If that doesn't work, consider re-entering the SNMP credentials on the device.

View solution in original post

6 Replies 6

Joe Clarke
Cisco Employee
Cisco Employee

For inventory, you need to make sure the SNMP read-only credential is correct in DCR.  You can look at the IC_Server.log for errors related to inventory collection.  Typically, errors about failed authentication are correct.  You might try starting a sniffer trace filtering on all SNMP traffic (udp/161) to the device.  then go to RME > Devices > Inventory > Inventory Jobs, and create a new job to collect inventory from that one failing device.  When it fails, check the sniffer trace to see what SNMP credentials are being used, and what the device is reporting.

Thanks Much Joe! I got this error on 2960 switches failed inventory on IC_server log "snmpRspWrongDigests" what does it mean? I have checked both snmp credential currently we have snmp v3 RW and snmp v2 RO both credential is correct. also if indeed some kind of string error " why I can snmpwalk on both V3 and V2 from device center/ and from the server itself with out a problem. I am a little bit confused. Thanks!!  I have attached the  log after I run inventory  on one of failed device.

This error means the SNMPv3 credentials in DCR are wrong for this device.  Check that you have entered the proper SNMPv3 username, password, and authentication algorithm (either MD5 or SHA-1).  The SNMP Walk from Device Center may work because you're entering the proper v3 credentials there.  Note: if you have SNMPv3 credentials configured in DCR, then the v2c credentials will not be used.

our credentail is entered as default, which applying for all devices. we are not enter credential per devices. but the defult applying for all, also I am using this defualt credential from device center too. so do you think something wrong with the device itself instead of DCR? because this default credentail working without any problem at least 3000+ devices handfull of them have this problem. so I am suspected may be the key generated on this devices may be wrong? or you still suggesting me focus on DCR credentail. again thanks so much for your time.

A wrong digest message means that the polling manager is using the wrong SNMPv3 password, or the wrong digest algorithm.  Since you said you can SNMP Walk the device using SNMPv3 from Device Center, I suspected the DCR credentials are wrong.  Therefore, I suggest you re-enter them in DCR (and verify the authentication algorithm).  If that doesn't work, consider re-entering the SNMP credentials on the device.

Thank you so very much!