mac discovery problems on Procurve 2524

[Daniel Sanabria]

I'm implementing a two switch stack where each is going to handle 2 VLANs, each switch will have the same port assigment for redundancy purposes. Once I configured the swicthes I noticed that when I ping some of the servers sometimes I get replies and some times I dont. From one of the server's console connection if I ping or telnet any other server I always obtain a positive response.

I decided to check the MAC table and I noticed that not all the mac are listed and my ping to all those addresses are getting valid responses. If I check the status I can't see any alarms.

Why the swicthes are not discovering all mac connected?

SwitchC# show config

Startup configuration:

; J4813A Configuration Editor; Created on release #F.05.17

hostname "SwitchC"
snmp-server contact
snmp-server location
time timezone -360
time daylight-time-rule Continental-US-and-Canada
cdp run
mirror-port 7
ip default-gateway 10.162.60.11
snmp-server community "public" Unrestricted
vlan 1
name "OAM_LAN"
untagged 1-7,13-18,23-26
ip address 10.162.60.20 255.255.255.0
no untagged 8-12,19-22
exit
vlan 2
name "MSG_LAN"
untagged 8-12,19-22
ip address 102.183.33.24 255.255.255.0
exit
stack commander "Stack2"
stack auto-grab
stack member 1 mac-address 0014c2
no aaa port-access authenticator active

SwitchC# show mac

Status and Counters - Port Address Table

MAC Address Located on Port
------------- ---------------
000d60- 14
001279- 22
001279- 9
001279- 1
001279- 12
001279- 7
001279- 20
001279- 8
001279- 19
0014c2- 7
0014c2- 7
000d60- 14
001279- 22
001279- 9
001279- 1
001279- 12
001279- 7
001279- 20
001279- 8
001279- 19
0014c2- 7
0014c2- 7

SwitchC# menu
SwitchC 1-Jan-1990 0:04:29
===========================- TELNET - MANAGER MODE -============================
Stacking - Stack Access

SN MAC Address System Name Device Type Status
-- ------------- ---------------- ----------- -------------------------
0 0014c2-780a00 SwitchC HP 2524 Commander Up
1 0014c2-75af00 SwitchD HP 2524 Member Up



Actions-> Cancel eXecute Help

Access Stack Member
Use arrow keys to change field selection



SwitchD# show mac

Status and Counters - Port Address Table

MAC Address Located on Port
------------- ---------------
000d60- 7
001279- 22
001279- 9
001279- 1
001279- 12
001279- 3
001279- 20
001279- 21
001279- 8
001279- 19
001279- 1
001279- 12
0014c2- 7
0014c2- 7
000d60- 7
001279- 22
001279- 9
001279- 1
001279- 12
001279- 3
001279- 20
001279- 21
001279- 8
001279- 19
001279- 1
001279- 12
0014c2- 7
0014c2- 7

SwitchD# show config

Startup configuration:

; J4813A Configuration Editor; Created on release #F.05.17

hostname "SwitchD"
snmp-server contact
snmp-server location
time timezone -360
time daylight-time-rule Continental-US-and-Canada
cdp run
mirror-port 7
ip default-gateway 10.162.60.11
snmp-server community "public" Unrestricted
vlan 1
name "DEFAULT_VLAN"
untagged 2-7,13-18,23-26
ip address 10.162.60.21 255.255.255.0
no untagged 1,8-12,19-22
exit
vlan 2
name "MSG_LAN"
untagged 1,8-12,19-22
ip address 102.183.33.25 255.255.255.0
exit
stack join 0014c27
no aaa port-access authenticator active

SwitchD#

[Matt Hobbs]

Are these switches connecting to a router further up? If yes, then your problem will be because you are connecting 2 uplinks to the upstream router. The solution is to use only one port and tag it in both VLANs.

There is an explanation on this here:

http://www.hp.com/rnd/support/faqs/23xx_25xx.htm#question27

[Daniel Sanabria]

Matt,

The siwtches are not connected to any router yet. I want to make sure first that communication between the servers and from the switches to any server is ok.

[Mohieddin Kharnoub]

Hi

- If the Servers located on the same Vlan and Same switch, then there should be no problem.

-If the Servers on the same Vlan but on different switches, then the uplink between switches should be tagged to vlan2 and untagged to vlan1.

-If the servers located on different vlans and whatever the switch, then your ping will go to your default gateway: 10.162.60.11, which is the router, so if this router is connected and configured to route between the 2 vlans, then everything should be ok.

-Try to upgrade the firmware for 2525 even its not your problem:
http://www.hp.com/rnd/software/j4812555.htm

Good Luck !!!

[Daniel Sanabria]

The servers are connected directly to the switches.

I just noticed that the problem only appears in the default LAN configured in the switch.

Everything is ok for vlan 2. The mac for these vlan are recorded ok and no problem when I ping to any server connected in this VLAN.