Wired Intelligent Edge

Hello all,

 I'm working in a big building with several associations.

We have create several VLAN on HP switch ( HP 2530-48G Switch)

    vlan 1, default VLAN  : for one association who use Windows 2016 server DC with DHCP and DNS, an exchange server and  a terminal server, video server...

    VLAN 100 for the wifi : zyxel controller with lots of Acess point.   ( NXC5200 )

    VLAN 101 for one association who use only google apps to work.

    VLAN 102 for horeca.

    VLAN 103 for an exhibition.

    Evrything was working fine till we had to work on the electricity, has we are still waiting for the UPS ( new server) I had to power off the physical server who is hosting all virtual server on VLAN 1.

    Since that shutdown, I have some PC who are randomly getting their DHCP request answered by the Zyxel controller.

    Sometimes is the IP, some times they get the correct IP by our DC but get the DNS by the zyxel.

    Running on the PC ipconfig /release   /flushdns  /renew and they get the IP by the correct DC.

    our DHCP lease is 12 hours normally but the  probleme is happening several times a day on the same PC.

    our subnet is 192.168.123.x 2   /24

    The Zyxell subnet is 10.100.1.x /16

    I know That DHCP request are send in broadcast but how comes that this request jump throught different vlan ?

    I was thinking could be an mistake on the switch configuration and try to add IP helper-address on vlan 1 but I get invalid input helper-address when I enter the command ... ???

    We have trunk between the different VLAN but this problem never occur before.

    I have shutdown the Zyxell and restart the DC, restart the DHCP and the DNS.

    restart all switch.

    The issue is going on since 5 days with some PC randomly getting the bad Ip adress or just the bad DNS adress.

    I have fix DNS server in some PC who where loosing the adress every 15 minutes but this is not the solution.

    I'm asking myselfwhere is the mistake in my config on switch ?

     

    switch master 02(config)# show config


    Startup configuration: 106

    ; J9775A Configuration Editor; Created on release #YA.15.12.0010
    ; Ver #04:01.ff.37.27:ea
    hostname "switch master 02"
    fault-finder bad-driver sensitivity high
    fault-finder bad-transceiver sensitivity high
    fault-finder bad-cable sensitivity high
    fault-finder too-long-cable sensitivity high
    fault-finder over-bandwidth sensitivity high
    fault-finder broadcast-storm sensitivity high
    fault-finder loss-of-link sensitivity high
    fault-finder duplex-mismatch-hdx sensitivity high
    fault-finder duplex-mismatch-fdx sensitivity high
    fault-finder link-flap sensitivity high
    trunk 25-26 trk1 lacp
    trunk 27-28 trk2 lacp
    trunk 29-30 trk3 lacp
    trunk 31-32 trk4 lacp
    trunk 33-34 trk5 lacp
    trunk 35-36 trk7 lacp
    trunk 37-38 trk8 lacp
    trunk 39-40 trk9 lacp
    trunk 47-48 trk10 lacp
    stack join 7446a0-e415c0
    ip default-gateway 192.168.123.254
    interface 1
       flow-control
       exit
    interface 2
       flow-control
       exit
    interface 3
       flow-control
       exit
    interface 4
       flow-control
       exit
    interface 5
       flow-control
       exit
    interface 6
       flow-control
       exit
    interface 7
       flow-control
       exit
    interface 8
       flow-control
       exit
    interface 25
       name "Sauveniere_1"
       exit
    interface 26
       name "Sauvienere_1"
       exit
    interface 27
       flow-control
       name "Sauveniere_2"
       exit
    interface 28
       flow-control
       name "Sauveniere_2"
       exit
    interface 29
       flow-control
       name "Sauveniere_3"
       exit

    interface 30
       flow-control
       name "Sauveniere_3"
       exit
    interface 31
       flow-control
       name "Sauveniere_4"
       exit
    interface 32
       flow-control
       name "Sauveniere_4"
       exit
    interface 33
       flow-control
       name "Sauveniere_5"
       exit
    interface 34
       flow-control
       name "Sauveniere_5"
       exit
    interface 35
       flow-control
       name "Sauveniere_7"

       exit
    interface 36
       flow-control
       name "Sauveniere_7"
       exit
    interface 37
       flow-control
       name "neujean_mediatheque"
       exit
    interface 38
       flow-control
       name "neujean_mediatheque"
       exit
    interface 39
       flow-control
       name "Neujean"
       exit
    interface 40
       flow-control
       name "Neujean"
       exit
    interface 45
       flow-control
       lacp active
       exit
    interface 46
       flow-control
       lacp active
       exit
    interface 47
       flow-control
       name "LACP MASTER"
       exit
    interface 48
       flow-control
       name "LACP MASTER"
       exit
    snmp-server community "public" unrestricted
    snmp-server location "salle serveur 4 eme"
    vlan 1
       name "DEFAULT_VLAN"
       untagged 1-24,41-46,49-52,Trk1-Trk5,Trk7-Trk10
       ip address 192.168.123.252 255.255.255.0
       exit
    vlan 100
       name "Vlan_mnema"
       tagged 1-24,41-46,49-52,Trk1-Trk5,Trk7-Trk10
       no ip address
       exit
    vlan 101
       name "Vlan_101_wifi"
       tagged 1-24,41-46,49-52,Trk1-Trk5,Trk7-Trk10
       no ip address
       exit
    vlan 102
       name "Vlan_102_horeca"
       tagged 1-24,41-46,49-52,Trk1-Trk5,Trk7-Trk10
       no ip address
       exit
    spanning-tree
    spanning-tree Trk1 priority 4
    spanning-tree Trk2 priority 4
    spanning-tree Trk3 priority 4
    spanning-tree Trk4 priority 4
    spanning-tree Trk5 priority 4
    spanning-tree Trk7 priority 4
    spanning-tree Trk8 priority 4
    spanning-tree Trk9 priority 4
    spanning-tree Trk10 priority 4
    password manager

    switch master 02(config)#

    Thanks.

    Kind regards,

    Elisa

 

Just to be clear: the hosts that are having the issue are in VLAN1 right?
Putting an ip-helper on the VLAN1 interface won't stop the hosts from broadcasting their request, it will just duplicate each request.

It sounds to me like your Zyxel config needs fixing - is it supposed be running a DHCP server in VLAN1?

EDIT:

Nevermind, I thought you had to shut down the switches, not just the server. 

You need a DHCP relay when you want to forward DHCP discoveries through layer 3. It won't help you when two DHCP servers receive the packets by broadcast.

My guess is that you didn't saved the config on some switch and the loss of electricity deleted the running config. And now the VLAN 1 traffic is untagged to another interface in VLAN 100. 

That way the DHCP is answered by the faster server, and it would look like it is seemingly random.

You can test if the VLANs are connected by configuring a static IP address of a VLAN 100 client on a client in VLAN1. When you can reach the Zyxel router that way, the switch config is faulty somewhere.