HP 2915-8G management VLAN

[manick]

Hi all,

 

This may well be a lack of understanding on my part but....

 

I have an 8 port 2915 switch which is connected to our core ProCurve switch via RJ45s. We have several VLANs, one of which is our management VLAN. Ports 9-10 are enabled and they are untagged for the management VLAN as is the port on the switch to which it is connected, however I am unable to ping or connect to the switch in any way shape or form. That is however until I connect to port 7 (which has also been untagged for the management VLAN) then I can ping telnet etc. If I then connect a second cable to port 10 and unplug port 7 I can carry on as normal (port 10 remains pingable etc.) however, if I unplug port 10 I am unable to reconnect to it if I plug it in again.

 

There are no GBIC modules installed.

 

Any ideas?

 

Many thanks

 

Nick

[twin--turbo]

Is the vlan tagged between switches?

[manick]

HI TT,

 

It's untagged. When the cable is plugged into port 7 (which has exactly the same tagging, untagging etc.) it works fine. Ports 9 and 10 won't work unless I have port 7 plugged in first then plug into port 9 or 10 (whilst 7 still plugged in) and then unplug 7 (leaving 9 or 10 in).

 

Nick

[twin--turbo]

Can you post up the config from the CLI?

 

Sounds like the vlan is only being brought up when activity is seen on 7

[manick]

Here you go with a couple of changes

 

Running configuration:

 

; J9562A Configuration Editor; Created on release #A.15.12.0010

; Ver #04:01.ff.36.27.fb

hostname "Cashless "

max-vlans 30

sntp server priority 1 xx.xx.xx.xx

sntp server priority 2 xx.xx.xx.xx

no stack

time daylight-time-rule western-europe

ip default-gateway CORE SWITCH gateway IP

snmp-server community "dhgreadwrite" unrestricted

snmp-server community "dhgread"

snmp-server host EDGE SWITCH IP community "dhgreadwrite" trap-level critical

snmp-server host EDGE SWITCH IP community "public"

vlan 1

name "Default"

no untagged 1-8

untagged 9-10

no ip address

exit

vlan 50

name "Voice"

tagged 1-10

ip address xx.xx.99.53 255.255.255.0

qos priority 7

exit

vlan 100

name "dhg_curric"

untagged 1-8

tagged 9-10

ip address xx.xx.202.53 255.255.252.0

ip igmp

exit

vlan 105

name "dhg_stubyod"

tagged 9-10

no ip address

ip igmp

exit

vlan 110

name "dhg_stabyod"

tagged 9-10

no ip address

ip igmp

exit

vlan 115

name "dhg_wcurric"

tagged 9-10

no ip address

ip igmp

exit

vlan 120

name "dhg_wunman"

tagged 9-10

no ip address

ip igmp

exit

vlan 125

name "dhg_apple"

tagged 9-10

no ip address

ip igmp

exit

vlan 130

name "dhg_wguest"

tagged 9-10

no ip address

ip igmp

exit

vlan 135

name "dhg_comm"

tagged 9-10

no ip address

ip igmp

exit

vlan 140

name "dhg_mgmt"

tagged 7,9-10

ip address xx.xx.34.31 255.255.255.0

ip igmp

exit

vlan 145

name "dhg_ap"

tagged 9-10

no ip address

ip igmp

exit

vlan 150

name "dhg_ereg"

tagged 9-10

no ip address

ip igmp

exit

vlan 155

name "dhg_cctv"

tagged 9-10

no ip address

ip igmp

exit

vlan 160

name "dhg_ptrs"

tagged 9-10

no ip address

ip igmp

exit

vlan 165

name "dhg_untwrd"

tagged 9-10

no ip address

ip igmp

exit

vlan 170

name "dhg_iptv"

tagged 9-10

no ip address

ip igmp

exit

vlan 175

name "dhg_access"

tagged 9-10

no ip address

ip igmp

exit

vlan 180

name "dhg_bms"

tagged 9-10

no ip address

ip igmp

exit

vlan 185

name "dhg_cashless"

tagged 1-10

no ip address

ip igmp

exit

spanning-tree

spanning-tree force-version rstp-operation

password manager

[twin--turbo]

why are 9-10 tagged on every vlan ??

 

what port is the uplink to the core ???

 

TT

[manick]

That's historical, before my time, not me guvnor!, but for the purposes of connecting to the IP of the management Vlan it shouldn't make any difference should it?. Re the uplink, at the moment 7 but I'm trying to get 9 or 10, the core switch port it's connected to has the management VLAN tagged.

 

Cheers

 

Nick

[twin--turbo]

vlan 1

name "Default"

no untagged 1-8

untagged 9-10

no ip address

exit

vlan 50

name "Voice"

tagged 1-10

ip address xx.xx.99.53 255.255.255.0

qos priority 7

exit

vlan 100

name "dhg_curric"

untagged 1-8

tagged 9-10

ip address xx.xx.202.53 255.255.252.0

ip igmp

exit

 

 

 

port 7,8,9,10 are not untagged on any access port.

 

 

 

there are no untagged ports on the mgmt vlan

 

vlan 140

name "dhg_mgmt"

tagged 7,9-10

ip address xx.xx.34.31 255.255.255.0

ip igmp

exit

vlan

 

 

TT

[manick]

I was under the impression as long as 9 or 10 are tagged on the mgmt-vlan and the port (to which it is attached) on the core switch is tagged I should be able to reach the IP (as with port 7). Why would I need to untag the ports?

 

Thanks

 

Nick

 

Sorry just noticed I said ports 7,9 and 10 were untagged for the VLAN in an earlier post. it should have read tagged obviously.

Edited March 9, 2016 by manick

[twin--turbo]

Which is the uplink port?

 

You would only TAG a port on a trunk so where are 8,9,10

[manick]

Isn't trunking aggregating ports in HP speak? the uplink port at present is port 7 as I can get no traffic through ports 9 or 10 (unless I've connected to port 7 first). All my management traffic is put through vlan 140 and so at present ports 7, 9 and 10 are tagged for this VLAN. This config is working for all other switches on my network.

 

I am very obviously missing something here, but don't know what.

 

cheers

 

Nick

[twin--turbo]

I am using "Trunk" in the sense that it's got multiple vlans tagged and not an "Access" port

 

I am unclear of what you are trying to achieve from your initial description.

 

It looks like your trying to use 9-10 for the uplink given that it's tagged on all but the default vlan.

 

TT

[manick]

HI TT,

 

OK, For this particular switch, HP 2915-8G, all I want to do at present is connect it via port 9 or 10 to our core switch and be able to putty to it using an IP address I have given it which is part of our mgmt vlan (140). I will be using this switch to connect some cashless tills and a cashless server (cashless-VLAN, 185). All is working fine with the exception of connecting via ports 9 or 10. I can connect using either a curriculum vlan IP or mgmt-vlan IP via port 7, but not port 9 or 10. As the switches are 802.1Q aware surely for the mgmt-vlan the port needs tagging (this is proved as I can connect using port 7), so at the moment it's looking like the issue is specific to ports 9 and 10 even though they are enabled and no GBIC module is installed, so theoretically should work.

 

Thanks for the input BTW!

 

Nick

[Blue_Cookeh]

Can you post the config of your core switch (or just the VLANs and ports concerned)?

 

You don't necessarily only tag ports on "trunks" (Cisco speak, a HP Trunk are bonded ports). So long as ports 9-10 on your smaller switch, and the ports you're connecting to at your core contain the VLANs on their interface config you should be good to go, it could be your routing/ACLs are screwed up somewhere instead.

Edited March 10, 2016 by Blue_Cookeh

[twin--turbo]

Have you just got the one cable to the core in at any one time?

 

Is vlan 140 tagged on the link from the core?

 

 

Output.

 

 

Sh vlan 145

Sh int 9

 

Try on both switches

 

Clear arp

 

Also try shutting dow the ports at each end then bring them back up.

[manick]

Hi BC,

 

Here is the core switch config, irrelevant VLANs have been taken out. The 8 port switch is connected to H17 on the core switch!

 

; J8698A Configuration Editor; Created on release #K.15.12.0012

; Ver #05:08.41.ff.9f.eb:22

hostname "core1"

module 1 type j9534a

module 2 type j9536a

module 3 type j9534a

module 4 type j9549a

module 5 type j9534a

module 6 type j9549a

module 7 type j9549a

module 8 type j9549a

module 11 type j9309a

timesync sntp

sntp unicast

sntp 30

sntp server priority 1 xx.xx.xx.xx

sntp server priority 2 xx.xx.xx.xx

ip route 0.0.0.0 0.0.0.0 xx.xx.xx.xx

ip routing

ip multicast-routing

snmp-server community "dhgreadwrite" unrestricted

snmp-server community "dhgread"

snmp-server host EDGE SWITCH IP community "XXXXXXXXXXXXXXX"

snmp-server contact "someone’s name" location "CR3 Cabinet"

router pim

enable

bsr-candidate priority 255

exit

vlan 1

name "vlan-1"

no untagged A1-A24,B1-B20,C1-C24,D1-D19,E1-E24,F1-F20,H1-H19

untagged B21-B22,D20-D24,F21-F24,G1-G24,H20-H24,K1-K4

ip address xx.xx.xx.50 255.255.255.0

exit

vlan 50

name "Voice"

tagged A1-A24,B1-B22,C1-C24,D1-D24,E1-E24,F1-F24,H1-H24,K1-K4

ip address xx.xx.99.53 255.255.255.0

qos priority 7

exit

vlan 100

name "dhg_curric"

untagged A1-A23,B1-B20,C1-C24,D1-D19,E1-E24,F3-F20,H1,H3-H16,H18-H19

tagged B21-B22,D21-D24,F1-F2,F21-F24,H21-H24,K1-K4

ip address xx.xx.xx.254 255.255.252.0

ip igmp

exit

vlan 140

name "dhg_mgmt"

tagged B21-B22,D21-D24,F1-F2,F21-F24,H17-H18,H21-H24,K1-K4

ip address xx.xx.34.31 255.255.255.0

ip igmp

exit

vlan 300

name "dhg_wanedge"

untagged A24,H2

ip address EDGE SWITCH IP 255.255.255.128

ip igmp

ip pim-sparse

ip-addr any

exit

exit

spanning-tree

spanning-tree priority 0 force-version rstp-operation

no autorun

no dhcp config-file-update

no dhcp image-file-update

password manager

Edited March 10, 2016 by manick
port info

[manick]

Just spotted that on vlan 1 on the core switch that H17 is 'no untagged', does it need to be 'untagged'????

 

Having said that if I can connect via port 7............

Edited March 10, 2016 by manick

[manick]

They are tagged at both ends (port 9 and port h17) for vlan 140 TT

[twin--turbo]

Just spotted that on vlan 1 on the core switch that H17 is 'no untagged', does it need to be 'untagged'????

 

No it means it should be untagged somewhere else... But I can't see where.

 

 

I hate the way Procurve handles vlan config. It is so messy.

 

TT

[manick]

Morning!

 

It's untagged on VLAN 100.

 

Nick