snmp missing real interfaces.


#1

1.1.7 on NTT’s cloud (because they rendered the insanely overpriced vyatta unusable.)

When I have a VRRP group, zabbix (beloved of Japan because “reasons”) only sees the vrrp interfaces like:

eth0v10, eth0v20, eth0v30

But not the underlying interface eth0 that is pushing all the traffic, including l2tp which is very busy.

Do I need to make some magical setting somewhere to enable this?
Is this a feature?


show snmp mib ifmib
eth0: ifIndex = 2
ifAlias = Outside
eth0v10@eth0: ifIndex = 5
eth0v20@eth0: ifIndex = 6
eth0v30@eth0: ifIndex = 9
eth1: ifIndex = 3
ifAlias = DMZ
eth1v11@eth1: ifIndex = 7
eth2: ifIndex = 4
ifAlias = Service
eth2v12@eth2: ifIndex = 8
l2tp0: ifIndex = 48
ifAlias = User: bunny (10.41.10.240)
lo: ifIndex = 1

For some reason zabbix isn’t seeing them, so I think they are really there and this is just a “pilot should not be allowed out of her wheelchair” type problem.


#2

From your output, wouldn’t do ifIndex=2 do the trick?
Play around with tool like snmpwalk to find out if this is Zabbix or Vyos issue


#3

Good idea.

I suspect the former, but this is only my second week of using either.
It would be nice if the SNMP would tell me how busy vyos was too. eg, CPU/Memory/Core Meltdown Imminent. Any VYOSNMP heroes care to give me enough rope to bungee jump with?


IF-MIB::ifDescr.1 = STRING: lo
IF-MIB::ifDescr.2 = STRING: Red Hat, Inc Device 0001
IF-MIB::ifDescr.3 = STRING: Red Hat, Inc Device 0001
IF-MIB::ifDescr.4 = STRING: Red Hat, Inc Device 0001
IF-MIB::ifDescr.5 = STRING: eth0v10
IF-MIB::ifDescr.6 = STRING: eth0v20
IF-MIB::ifDescr.7 = STRING: eth1v11
IF-MIB::ifDescr.8 = STRING: eth2v12
IF-MIB::ifDescr.9 = STRING: eth0v30
IF-MIB::ifDescr.57 = STRING: eth0v40
IF-MIB::ifDescr.65 = STRING: ppp0
IF-MIB::ifDescr.69 = STRING: Red Hat, Inc Device 0001
IF-MIB::ifDescr.70 = STRING: eth3v13
IF-MIB::ifDescr.73 = STRING: l2tp2

I guess zabby doesn’t like all these devices with the same descr. Is there anyway to change that in vyos?


#4

Now I get different results with snmp walk on the the 2 VRRP vyos 1.1.7 boxes:

primary:
snmpwalk -v 2c -c public 10.41.0.3 . > out2.txt
Error: OID not increasing: IP-FORWARD-MIB::ipCidrRouteDest.10.40.12.0.0.255.255.255.0.0.0.0.0

= IP-FORWARD-MIB::ipCidrRouteDest.10.40.12.0.0.255.255.255.0.0.0.0.0

secondary:
snmpwalk -v 2c -c public 10.41.0.4 . > out3.txt
(not a squeak)

-rw-r–r-- 1 peter users 43443 Jun 23 11:21 out2.txt
-rw-r–r-- 1 peter users 556570 Jun 23 11:21 out3.txt

This looks like an snmpwalk bailout, but should the OID do this in the case of multiple routes:

vyos@vyos:/etc/snmp$ show ip route
Codes: K - kernel route, C - connected, S - static, R - RIP, O - OSPF,
I - ISIS, B - BGP, > - selected route, * - FIB route

S>* 0.0.0.0/0 [1/0] via 10.0.0.33, eth0
R>* 10.40.10.0/24 [120/6] via 10.40.12.5, eth3, 20:27:01
C * 10.40.12.0/24 is directly connected, eth3v13
C>* 10.40.12.0/24 is directly connected, eth3
C * 10.41.0.0/24 is directly connected, eth1v11
C>* 10.41.0.0/24 is directly connected, eth1
C * 10.41.10.0/24 is directly connected, eth2v12
C>* 10.41.10.0/24 is directly connected, eth2
C>* 10.41.10.240/32 is directly connected, l2tp0
C>* 10.41.10.241/32 is directly connected, l2tp1
C>* 10.41.10.242/32 is directly connected, l2tp2
C>* 127.0.0.0/8 is directly connected, lo


#5

On my test-router, I also have multiple routes double:

O 10.255.255.0/30 [110/10] is directly connected, vti0, 05w1d06h
C>* 10.255.255.0/30 is directly connected, vti0
O 10.255.255.4/30 [110/15] is directly connected, vti1, 05w1d06h
C>* 10.255.255.4/30 is directly connected, vti1

However, only routes with > end up in table and are shown in SNMP:

OID=.1.3.6.1.2.1.4.24.4.1.1.10.255.255.0.252.255.255.255.0.0.0.0.0, Type=IpAddress, Value=10.255.255.0
OID=.1.3.6.1.2.1.4.24.4.1.1.10.255.255.4.252.255.255.255.0.0.0.0.0, Type=IpAddress, Value=10.255.255.4
OID=.1.3.6.1.2.1.4.24.4.1.1.10.255.255.8.252.255.255.255.0.0.0.0.0, Type=IpAddress, Value=10.255.255.8