Discovery error with FindIT Probe on SG350X

[Christian Woegerbauer]

Hi All!

We got a problem with discovering a SG350X (stack with 4 switches) for the Cisco Business Dashboard.

The FindIT Network Probe is activated on the switch itself. When we start the refresh in inventory it find itself as a device (switch, model, version, SN and MAC). But no hostname, no IP address and not that this switch consists of 4 switches (see screenshot 14-sw01 attached).

A non Cisco WiFi-AP (Ubiquity) is correctly discovered.

We got the same setup on a different site of the customer, here the FindIT Probe on the switch discovers itself correctly (see screenshot 23-sw01).

Both switches-stacks run firmware 2.5.0.92 and report to the Cisco Business Dashboard.

Credentials for discovery should be correct as there are no needed device credentials under "Administration - Device Credentials" in the web gui of the switch's FindIT Network Probe.

Thanks in advance,

Chris

[David Harper]

Hi Chris,

What you describe is certainly not the expected behaviour.  How long did you wait after enabling the probe before capturing the screenshot?  It will take the probe a couple of minutes to start up and fully discover the network, and the screenshot is typical for when the probe is part way through the discovery process.  But I amd only talking maybe 5-10 minutes at most for a complex network, so unless you captured the screenshot very soon after enabling the probe, then there is probably a different explanation.

Would you be able to PM the switch config to me, along with a description of the real network topology?  There might be something unusual in your setup that is causing the problem.

On another note, I would also recommend upgrading to the 2.5.5 version of firmware for the switch.  It shouldn't have any bearing on this problem, but it does provide a much tighter integration with Cisco Business Dashboard than the older versions of firmware.

Cheers,

Dave.

[Christian Woegerbauer]

Dear David!

Even hours of starting a manual inventory, the status of the faulty discovery of the switch is the same.

I'll send you the configuration of it via PM.

The real network topology is the following:

Site 14 (where the discovery problem occurs):

2 VLANS (Office and Management)

Switches and devices which should be discovered and monitored are in the Management VLAN.

Switchstack of 4 SG350X where FindIT NW-Probe is activated.

Site 23 (where the discovery runs correctly):

2 VLANS (Office and Management)

Switches and devices which should be discovered and monitored are in the Management VLAN.

Switchstack of 4 SG350X where FindIT NW-Probe is activated.

Virtual Machine with Cisco Business Dashboard resides on this site in the Management VLAN. Local probe is disabled.

Both sites are connected via a VPN-Tunnel, no access-restriction between them. And both sites have the same layout and design.

As the discovery in Site23 runs correctly and the WiFi-AP on Site14 is displayed correctly in the CBD, i don't think its a problem with the network topology or the VPN-connection between the 2 sites. Especially because even in the WEB-GUI of the Site14-Switch it is not discovered correctly. Same layout and design at Site23, here everything works as expected.

Thanks in advance and BR,

Christian

[David Harper]

Hi Christian,

Looking at the config you sent through, I think you are running into a bug that is triggered when VLAN 1 is disabled or does not have an IP address. I suspect the switch at the other site is using different vlan IDs and is using VLAN 1.  The bug was fixed in switch firmware 2.5.5 so I would recommend upgrading.  If you need to keep on the same firmware, then a workaround is to bring VLAN 1 up and give it an IP address, and then discovery should start working.

Cheers,

Dave.

[Christian Woegerbauer]

Hello Dave!

VLAN 1 on the 14-sw01 was active but had no IP as you saw in the configuration. I added an IP Address to the VLAN and startet the discovery again. Now the Inventory shows the IP of the switch, but it's still not recognised as a 4 switch stack nor the hostname.

I'll have to talk to the customer regarding a firmware upgrade. Do you have a bug ID for the discovery-error with non active VLAN 1 you mentioned?

Thank you and Best Regards,

Christian

[David Harper]

Is the VLAN 1 interface active?  If you look in the GUI at IP Configuration > IPv4 Management and Interfaces > IPv4 interface or 'show ip interface' in the CLI, does the VLAN 1 interface show as valid?  Normally the interface will be down unless there is at least one port in the vlan (access or trunk) that is up.

There is not a publicly visible bug ID for this problem as it was identified only during internal testing, so the issue was tracked and fixed using the internal development tools only.

Cheers,

Dave.

[Christian Woegerbauer]

Hello David!

Yes, it's up:

IP Address I/F I/F Status Type Directed Prec Redirect Status

10.10.102.21/24 vlan 1 UP/UP Static disable No enable Valid

We'll upgrade the firmware to the latest and check the discovery afterwares.

Best Regards,

Christian