what causes a topology change in rapid-pvst mode

[seanxiao]

hi all,

Thanks for your reply!

physical topology is very simple: core switch stack (2x WS-C6506-E) connects to access switch via port-channel, no more redundant link. I see lots of topology change logged from core switch, for example:

Number of topology changes 1184 last change occurred 11:26:37 ago
from Port-channel17

then I go to the access switch at the other end of port-channel17:

accessswitch#show spanning-tree detail | in ieee|from|occur|is exec
Number of topology changes 2 last change occurred 31w6d ago
from AppGigabitEthernet1/0/1
Number of topology changes 7 last change occurred 5d12h ago
from Port-channel1
Number of topology changes 109 last change occurred 06:49:59 ago
from Port-channel1
Number of topology changes 1799 last change occurred 11:34:50 ago
from GigabitEthernet2/0/1

let's take Gi2/0/1 as an example. I show run interface Gi2/0/1 on accessswitch:

interface GigabitEthernet2/0/1
description MEP2 STN90M CPR317-0090
switchport access vlan 2524
switchport mode access
no snmp trap link-status
end

accessswitch#show run | section spanning-tree
spanning-tree mode rapid-pvst
spanning-tree extend system-id
spanning-tree uplinkfast
spanning-tree vlan 108,2524,2617 hello-time 1
spanning-tree vlan 108,2524,2617 forward-time 4
spanning-tree vlan 108,2524,2617 max-age 6

If I show logging on the access switch, i do see a lot of logs about interface up/down, so the high rate of topology change is caused by the access port going up and down? But in my mind there is some article/some one saying RSTP will recognize the access port as edge port automatically, thus it won't cause any topology change (like Martin's reply in this post). Anyway,  in my case, should I explicitly issue spanning-tree portfast under access port and enable spanning-tree portfast bpduguard default globally in the access switch? 

[MHM Cisco World]

Did you check PO by

Show port channel summary 

Share it here for both VSS abd access SW

MHM

[seanxiao]

thank you MHM,

from VSS:

RS001#show interfaces port-channel 17 summary

*: interface is up
IHQ: pkts in input hold queue IQD: pkts dropped from input queue
OHQ: pkts in output hold queue OQD: pkts dropped from output queue
RXBS: rx rate (bits/sec) RXPS: rx rate (pkts/sec)
TXBS: tx rate (bits/sec) TXPS: tx rate (pkts/sec)
TRTL: throttle count

Interface IHQ IQD OHQ OQD RXBS RXPS TXBS TXPS TRTL
-----------------------------------------------------------------------------------------------------------------
* Port-channel17 0 0 0 1027651 70000 57 105000 66 0

from ACcess switch:

7#show interfaces port-channel 1 summary

*: interface is up
IHQ: pkts in input hold queue IQD: pkts dropped from input queue
OHQ: pkts in output hold queue OQD: pkts dropped from output queue
RXBS: rx rate (bits/sec) RXPS: rx rate (pkts/sec)
TXBS: tx rate (bits/sec) TXPS: tx rate (pkts/sec)
TRTL: throttle count

Interface IHQ IQD OHQ OQD RXBS RXPS TXBS TXPS TRTL
-----------------------------------------------------------------------------------------------------------------
* Port-channel1 0 0 0 0 105000 60 65000 55 0

[MHM Cisco World]

 not show interface port channel summary 
show port channel summary 

MHM

[seanxiao]

sorry, no command like this in both switch, 

access is Cisco IOS XE Software, Version 17.06.05

vss is Cisco IOS Software, s2t54 Software (s2t54-ADVENTERPRISEK9-M), Version 15.5(1)SY10,

[MHM Cisco World]

Show etherchannel summary 

Cisco always swapping the name' some times port channel and some etherchannel 

MHM

[seanxiao]

access:

#show etherchannel summary
Flags: D - down P - bundled in port-channel
I - stand-alone s - suspended
H - Hot-standby (LACP only)
R - Layer3 S - Layer2
U - in use f - failed to allocate aggregator

M - not in use, minimum links not met
u - unsuitable for bundling
w - waiting to be aggregated
d - default port

A - formed by Auto LAG

Number of channel-groups in use: 1
Number of aggregators: 1

Group Port-channel Protocol Ports
------+-------------+-----------+-----------------------------------------------
1 Po1(SU) - Gi1/1/1(P) Gi2/1/1(P)

VSS:

KRS001#show etherchannel summary
Flags: D - down P - bundled in port-channel
I - stand-alone s - suspended
H - Hot-standby (LACP only)
R - Layer3 S - Layer2
U - in use N - not in use, no aggregation
f - failed to allocate aggregator

M - not in use, no aggregation due to minimum links not met
m - not in use, port not aggregated due to minimum links not met
u - unsuitable for bundling
d - default port

w - waiting to be aggregated
Number of channel-groups in use: 42
Number of aggregators: 42

Group Port-channel Protocol Ports
------+-------------+-----------+-----------------------------------------------
1 Po1(SU) - Gi1/3/1(P) Gi2/3/1(P)
2 Po2(SU) - Gi1/3/2(P) Gi2/3/2(P)
3 Po3(SU) - Gi1/3/3(P) Gi2/3/3(P)
4 Po4(SU) - Gi1/3/4(P) Gi2/3/4(P)
5 Po5(SU) - Gi1/3/5(P) Gi2/3/5(P)
6 Po6(SU) - Gi1/3/6(P) Gi2/3/6(P)
7 Po7(SU) - Gi1/3/7(P) Gi2/3/7(P)
8 Po8(SU) - Gi1/3/8(P) Gi2/3/8(P)
9 Po9(SU) - Gi1/3/9(P) Gi2/3/9(P)
10 Po10(SU) - Gi1/3/10(P) Gi2/3/10(P)
11 Po11(SU) - Gi1/3/11(P) Gi2/3/11(P)
12 Po12(SU) - Gi1/3/12(P) Gi2/3/12(P)
13 Po13(SU) - Gi1/3/13(P) Gi2/3/13(P)
14 Po14(SU) - Gi1/3/14(P) Gi2/3/14(P)
15 Po15(SU) - Gi1/3/15(P) Gi2/3/15(P)
16 Po16(SU) - Gi1/3/16(P) Gi2/3/16(P)
17 Po17(SU) - Gi1/3/17(P) Gi2/3/17(P)
18 Po18(SU) - Gi1/3/18(P) Gi2/3/18(P)
19 Po19(SU) - Gi1/4/33(P) Gi2/4/13(P)
60 Po60(SU) LACP Te1/6/29(P) Te2/6/28(P)
61 Po61(SU) LACP Te1/6/30(P) Te2/6/27(P)
111 Po111(RU) - Te1/5/4(P) Te1/5/5(P)
112 Po112(RU) - Te2/5/4(P) Te2/5/5(P)
201 Po201(SU) - Gi1/3/47(P) Gi1/3/48(P)
202 Po202(SD) - Gi1/3/45(D)
301 Po301(SU) - Gi1/4/47(P) Gi2/4/47(P)
302 Po302(SU) - Gi1/4/46(P) Gi2/4/46(P)
303 Po303(SU) - Gi1/4/45(P) Gi2/4/45(P)
304 Po304(SU) - Gi1/4/44(P) Gi2/4/44(P)
305 Po305(SU) - Gi1/4/43(P) Gi2/4/43(P)
309 Po309(SU) - Gi1/4/39(P) Gi2/4/39(P)
310 Po310(SU) - Gi1/4/38(P) Gi2/4/38(P)
311 Po311(SU) - Gi1/4/37(P) Gi2/4/37(P)
312 Po312(SU) - Gi1/4/36(P) Gi2/4/36(P)
313 Po313(SU) - Gi1/4/35(P) Gi2/4/35(P)
314 Po314(SU) - Gi1/4/34(P) Gi2/4/34(P)
500 Po500(SU) - Gi1/4/5(P) Gi1/4/6(P) Gi1/4/7(P)
Gi1/4/8(P) Gi2/4/5(P) Gi2/4/6(P)
Gi2/4/7(P) Gi2/4/8(P)
501 Po501(SU) - Gi1/4/9(P) Gi1/4/10(P) Gi1/4/11(P)
Gi1/4/12(P) Gi2/4/9(P) Gi2/4/10(P)
Gi2/4/11(P) Gi2/4/12(P)
502 Po502(SU) - Gi1/4/1(P) Gi1/4/2(P) Gi2/4/3(P)
Gi2/4/4(P)
503 Po503(SU) - Gi1/4/3(P) Gi1/4/4(P) Gi2/4/1(P)
Gi2/4/2(P)
504 Po504(SU) LACP Te1/6/1(P) Te1/6/2(P) Te2/6/29(P)
Te2/6/30(P)
505 Po505(SU) LACP Te1/6/31(P) Te1/6/32(P) Te2/6/31(P)

[MHM Cisco World]

Hi friend 

So etherchannel summary is ok the PO is SU and eack port member is P

Can you check the show spanning tree

See if PO appear or it port member appesr in STP.

Also I see you use spanning tree uplinkfast' I am not so sure but this feature is add only when we have two uplink one FWD and other BLK and it fast STP to make port FWD.

It can bug or not but you dont need this feature if you have one PO uplink.

Check this point 

MHM

[seanxiao]

yeah thanks MHM, I am having the same view as you, cisco already recommended, and "Cisco recommends that UplinkFast be enabled only for switches with blocked ports, typically at the access-layer. Do not use on switches without the implied topology knowledge of a alternative/backup root link typically to distribution and core switches in Cisco multilayer design.

so, in my case, what I should do is:

1.explicitly issue spanning-tree portfast under access port and enable spanning-tree portfast bpduguard default globally in the access switch. IN this way there won't be TC occur if an Access port down and up.

2. remove uplinkfast feature globally.

can you please advise.

[MHM Cisco World · Accepted solution]

1.explicitly issue spanning-tree portfast under access port and enable spanning-tree portfast bpduguard default globally in the access switch. IN this way there won't be TC occur if an Access port down and up.
Answer:- the portfast and BPDU guard is use only for port connect to host or server not port that interconnect two SW

2. remove uplinkfast feature globally.
Answer:- Yes remove this command and monitor your SW

MHM