Cisco 2960S - Stack Failover

bunce.jake · ‎04-12-2012

Hi,

I have a Cisco 2960S stack and I'd like to tune the timers so that packet loss is minimal if a switch fails.

To simulate a failed switch I have reloaded it while running a continuous ping to a management address on the stack's SVI:

Reply from 192.168.101.246: bytes=32 time=1ms TTL=255

Reply from 192.168.101.246: bytes=32 time=2ms TTL=255

Reply from 192.168.101.246: bytes=32 time<1ms TTL=255 !! -- This is when I issued the reload command for the master switch only

Request timed out.

Reply from 192.168.101.2: Destination host unreachable.

Request timed out.

Reply from 192.168.101.246: bytes=32 time<1ms TTL=255

The topology used to test this failover is as follows:

[Catalyst 2960S - 2 in Stack]====Cross stack Etherchannel=====[Single HP Switch 802.3ad Etherchannel (LACP Active)]

I've also tested via this topology:

[Catalyst 2960S - 2 in Stack]====Cross stack Etherchannel=====[PAN Firewall with portfast trunk static Etherchannel (Mode on)]

As I see the same results when a ping is directed through the single switch (switched), and through the firewall (routed), I'm inclined to think that this is due to the stack failover timers on the Catalyst 2960S.

Is it possible to change the stack failover timers (i.e hold down, failover etc), to speed up the failover process?

Regards,

Jacob

amonehin02 · ‎06-04-2012

Hi Jacob,

Are you connected to the second switch in the stack by console?

Do you see any messages to confirm that the second switch becomes the master?

Also, do you see the port channel up via console or on the other device?

Rgrds,

Dami

p.juarezponte · ‎09-11-2018

Hello,

This is an old post but I would like to know if this problem could be solved.

I am having this same problem.

We have an ip phone (it's not really that but it works like that) which has two interfaces.

One NIC connected to switch 1 and the other to the switch 2 (it's a 2960-X stack).

If I unplug one of the wires (the one connected to the master and then the one connected to the member switch), the call remains established and I can talk to the other.

We are making some high availability tests and when I reload the switch which is the master, the call is missed and I lose connectivity for a few seconds (about 40).

I changed spanning tree from pvst to rapid pvst, and I could reduce it to 8-10s.

It's a critical environment where no one call can be missed.

Has anyone could solve this?

This is the log:

Sep 11 09:03:38.603: %STACKMGR-4-SWITCH_REMOVED: Switch 1 has been REMOVED from the stack
Sep 11 09:03:38.610: license_init_config_queue: initializing the license_switch_boot_config_q
Sep 11 09:03:38.617: %STACKMGR-4-MASTER_ELECTED: Switch 2 has been elected as MASTER of the stack
Sep 11 09:03:38.806: %CFGMGR-6-APPLYING_RUNNING_CFG: as new master
Sep 11 09:03:38.903: %SSH-5-ENABLED: SSH 1.99 has been enabled
Sep 11 09:03:39.858: %STACKMGR-5-MASTER_READY: Master Switch 2 is READY
Sep 11 09:03:39.879: %COMMON_FIB-4-FIBIDB: No fibidb found for Virtual1 during address change
Sep 11 09:03:40.662: %LINK-5-CHANGED: Interface Vlan1, changed state to administratively down
Sep 11 09:03:40.662: %LINK-3-UPDOWN: Interface Vlan52, changed state to up
Sep 11 09:03:40.662: %LINK-5-CHANGED: Interface FastEthernet0, changed state to administratively down
Sep 11 09:03:41.665: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet0, changed state to down
Sep 11 09:03:41.895: %LINK-5-CHANGED: Interface GigabitEthernet2/0/26, changed state to administratively down
Sep 11 09:03:42.895: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet2/0/26, changed state to down
Sep 11 09:03:43.902: %LINK-3-UPDOWN: Interface GigabitEthernet2/0/26, changed state to up
Sep 11 09:03:48.718: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan52, changed state to up
Sep 11 09:03:49.683: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet2/0/26, changed state to up
Sep 11 09:03:50.885: %LINK-3-UPDOWN: Interface Port-channel2, changed state to up
Sep 11 09:03:51.888: %LINEPROTO-5-UPDOWN: Line protocol on Interface Port-channel2, changed state to up

NguyenHung2238 · ‎04-20-2021

Hi, Any solution for this case? I had same problem.