topic snmp in ha mode of smb spark 1500 in Spark Firewall (SMB)

snmp in ha mode of smb spark 1500

RioAung — Wed, 07 Sep 2022 02:22:38 GMT

Hi ,

I would like to ask about snmp configuration with prtg on quantum spark 1500 HA mode.

Please see below diagram. i can monitor of same network VIP ( HA IP -10.10.20.1) from my monitoring server.

but i cannot monitor to APP Firewall Virtual IP HA ( 10.10.10.2) . I also cannot ping to this 10.10.10.2 ip from monitoring server .But i can reachable and monitor the devices behind the App Fw ( 10.10.30.0/24) .

So let me know why monitoring workstation cannot ping or cannot monitor the VIP of APP firewall HA IP ?

i cannot monitor because of HA IP ?

can i add custom OID on prtg to monitor detail feature of member down/ha down ?

can i monitor the cpu/memory wiht PRTG ?

Re: snmp in ha mode of smb spark 1500

G_W_Albrecht — Wed, 07 Sep 2022 08:19:18 GMT

With SMB cluster, only the active node does communication - backup node only syncs with active, but does not communicate with the world at all !

Quantum Spark 1500, 1600 and 1800 Appliance Series R80.20.40 Locally Managed Administration Guide p.30:

In High Availability, only one gateway is active at a time. When there is a failover, the standby member becomes active. There is no load sharing between the members of the cluster.

All cluster configuration is done through the active member.

Re: snmp in ha mode of smb spark 1500

RioAung — Thu, 08 Sep 2022 04:03:41 GMT

Hi @G_W_Albrecht ,

As per your comments, should we use active node physical ip for snmp traffic ?

how to know which interface are using for snmp traffic ?

can we choose the interface for snmp monitor ?

Re: snmp in ha mode of smb spark 1500

G_W_Albrecht — Thu, 08 Sep 2022 10:01:15 GMT

There is a big difference between GAiA Cluster XL and SMB HA Clustering - the SMB standby node is completely transparent. You contact it once for FTW with its physical IP to define as standby, and all other config details will be synced from active node. During the existence of the cluster, connections are always made to the VIP - if active node fails, VIP will connect to old standby now active node. Same should be true for SNMP monitoring as the active node has the needed information (e.g. VPN is up).

See sk121096: How to configure a cluster between locally managed SMB appliances:

The secondary appliance must not have any configuration. The only configuration should be on the hardware level (if a LAN interface is configured on the Active member, then the same Port should be configured on the Standby member).