Discovery not working properly for clustered firewall since upgrade to 20.4 CU4
search cancel

Discovery not working properly for clustered firewall since upgrade to 20.4 CU4

book

Article ID: 254612

calendar_today

Updated On:

Products

DX Unified Infrastructure Management (Nimsoft / UIM) CA Unified Infrastructure Management On-Premise (Nimsoft / UIM) CA Unified Infrastructure Management SaaS (Nimsoft / UIM)

Issue/Introduction

Discovery in the OC 20.45 does not work at all for us since we upgraded. Also, we have manually added items into our SNMPCollector now that do not show up when we look in the OC under inventory. We also have some of those devices that show up in the Performance Reports Designer and some do not.

At this point, we can't use discovery properly at all in our test or prod environments.

MAIN Issue - Discovery of a firewall cluster shows two IP addresses, one for each node but as a single inventory item, rather than multiple nodes/devices in the SNMP Collector.

Related issues:

  • Cannot view manually-added SNMPC Devices in OC Inventory.
  • Cannot see metrics in Performance Reports Designer from discovered firewall nodes at all, in the Performance Reports Designer typing in the hostname there is no device name option to choose from.

Environment

  • Release: DX UIM 20.4
  • discovery_server 20.41, and 20.42

Cause

  • This appears to be a vendor issue (All SysName's had defaulted to XXXX due to cluster/SNMP).

Resolution

As the SysName/DisplayName was the same for all the cluster nodes and correlation was happening because of the name (XXX-XX-XXXXXXXXXXX-XXXX) of the cluster nodes, in the discovery_server.cfg, we added the excluded_values for  (XXX-XX-XXXXXXXXXXX-XXXX) in AllCorrelationNames section as shown below:

<AllCorrelationNames>
            included_target_properties = "any.CorrelationNames"
            included_source_properties = "any.CorrelationNames"
            type = Other
            <excluded_values>
             xxx-xx-xxx =  (XXX-XX-XXXXXXXXXXX-XXXX)
            </excluded_values>
<AllCorrelationNames>

After this change, and cold starting the discovery_server, (Deactivate-Activate), we were able to see the 4 cluster nodes in the OC (with the same Name but different IpAddress) and there were 4 different cs_ids generated, one for each cluster node.

Setup and metrics were then working fine.

Additional Information

Modified discovery.cfg sections (for exclusions)
 
<MacAddresses>
            included_target_properties = "any.PrimaryMacAddress,any.OtherMacAddresses"
            included_source_properties = "any.PrimaryMacAddress,any.OtherMacAddresses"
            type = Other
            <excluded_values>
                  ... (entries removed)
...
...
...
...
...
...
 
               XXX_XX_XXXXXXXXXXX_XX_XXX_XX_X = <mac_address>
               XXX_XX_XXXXXXXXXXX_XX_XXX_XX_X = <mac_address>
            </excluded_values>

and...
 
<AllCorrelationNames>
            included_target_properties = "any.CorrelationNames"
            included_source_properties = "any.CorrelationNames"
            type = Other
            <excluded_values>
               switch_short_name = switch
               switch_long_name = switch.*
               router_long_name = router.*
               router_short_name = router
               localhost = localhost*
               xxx_xx_xxx = XXX-XX-XXXXXXXXXXX-XXXX
            </excluded_values>
         </AllCorrelationNames>