Line protocol on Interface Vlan changed state to up

[CaptainPistache]

Hello,

Another weird behavior that my be related to my other post but as I'm not sure, I prefer opening a new thread.

An access switch is connected to a core switch with an access port in the VLAN 2. From the core switch logs, I can sometimes see the following behavior :

Oct 1 05:34:39 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up
Oct 1 05:35:25 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to down
Oct 1 05:35:26 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up
Oct 1 05:35:50 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to down
Oct 1 05:35:50 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up
Oct 1 05:36:44 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to down
Oct 1 05:36:44 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up

Oct 2 05:34:02 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to down
Oct 2 05:34:02 UTC: %HSRP-5-STATECHANGE: Vlan2 Grp 0 state Standby -> Init
Oct 2 05:34:02 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up
Oct 2 05:34:02 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to down
Oct 2 05:34:02 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up
Oct 2 05:34:47 UTC: %HSRP-5-STATECHANGE: Vlan2 Grp 0 state Speak -> Standby
Oct 2 05:35:13 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to down
Oct 2 05:35:13 UTC: %HSRP-5-STATECHANGE: Vlan2 Grp 0 state Standby -> Init
Oct 2 05:35:13 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up
Oct 2 05:35:36 UTC: %HSRP-5-STATECHANGE: Vlan2 Grp 0 state Speak -> Standby
Oct 2 05:39:27 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to down
Oct 2 05:39:27 UTC: %HSRP-5-STATECHANGE: Vlan2 Grp 0 state Standby -> Init
Oct 2 05:39:28 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up
Oct 2 05:40:10 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to down
Oct 2 05:40:10 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan2, changed state to up
Oct 2 05:40:33 UTC: %HSRP-5-STATECHANGE: Vlan2 Grp 0 state Speak -> Standby

It happens about 4-5 times a day without any visible impact on our production. From the access switch, the uplink is always up. And as you can see, the HSRP state is therefore impacted.

I have made the following checks :

- VLAN exists in the vlan db and is active

- The VLAN interface is not shut

- An access port is inside this VLAN and its status is always connected

- Another trunk port allowed this VLAN

- There is not entries regarding STP in the syslog

- There's no error (input, CRC, etc) when checking the interface counters

Why does the line protocol for this specific VLAN sometimes go up/down ?

Thank you !

[balaji.bandi]

Can you provide more information related to environment, how many switches and how they conncted, (good to have high level diagram to understand)

when HSRP failing - this could be low reasons.

1. Lost connect between L3 connectivity.

2. this could me while spanning-tree convergence time

3. or some link failures.

this can only confirm based on some more inputs.

[CaptainPistache]

Hi balaji,

Thanks for your reply.

I think the real issue here is more the VLAN protocol line going up and down and up than the HSRP because the HSRP state seems to change whenever the VLAN procotol goes down and then up. Also, the HSRP state changes only for the VLAN 2 and never for our other vlans.

Here is a conceptual diagram of the switches involved into this situation :

CORE SWITCH 1 [HSRP active] (3750X) <------ trunk link with many vlans ------> CORE SWITCH 2 [HSRP standby] (3750X) <--- mode access vlan 2 ---> Access switch (2960X)

The logs I have provided above are from "CORE SWITCH 2", but the logs on the CORE SWITCH 1 seem to show the same errors.

Does it help ?

Thank you.

[balaji.bandi]

As per my understanding

You have CORE - CORE connected

3750 switch connect to each Core for the uplink

2960 access switch connected to 3750 ? is this correct

how is these connected L2 ? from 2960 to 3750 to Core ?

or L3 ?

[CaptainPistache]

---

@balaji.bandi wrote:

3750 switch connect to each Core for the uplink

2960 access switch connected to 3750 ? is this correct

Yes, this is correct.

---

@balaji.bandi wrote:

how is these connected L2 ? from 2960 to 3750 to Core ?

or L3 ?

---

Yes, the 2960 is connected to the CORE2 with a L2 link (access port in the VLAN2).

[balaji.bandi]

Can you post the configuration and spanning-tree config ? who is the root bridge here ?

[CaptainPistache]

Hi,

Sorry for the delay of my reply.

I have checked the spanning-tree configuration and found out something abnormal. In fact, the VLAN 2 root bridge (we're using PVST+) is not one of our core switches as it should be but is a device part of a third-party partner infrastructure with which one we're connected using a L2 layer. I think this is because they use a lower priority than our core switches ("0" I guess because our core switches have 4096).

So I have a good intuition that the issues comes from here. Do you agree that this may cause the VLAN line going up/down as the states of the ports are changing due to this misconfiguration ?

Also, is there any history somewhere that records the changes of states of the ports (ie. FWD to BLK) ?

[balaji.bandi]

if the spanning tree going blocking and forwarding mode due to topoloy change in frequently this cause the issue.

To sugget better, we would like to see your configuraiton and topology ?

[Giovanni Alexander Molina]

Hi, can you provide the configs of 3 switches?