Can Spanning Tree sometimes fail?

[abc1235]

Hallo,

My colleagues were changing some cables in a customer's building and accidentally someone connected 2 access switches. Of course this created a loop and the whole network in this building went down. After a while, the network came back on and so I had to check if STP was configured. All the ports on all access switches have STp configured which left me wondering what else could have caused this problem. Could anyone have another explanation for this scenario? Thanks

[MHM Cisco World]

there are two Loop 
spanning tree permanent loop

Spanning tree temporary loop <<- you face this since the Network work again normal, 

the spanning tree need time to detect loop and BLK port cause the loop. 
the dangerous loop is permanent. 

[abc1235]

So does it happen that sometimes STP takes time to detect a loop but when it does, it does a recovery? Is that correct?

[MHM Cisco World]

Yes correct
there is temporary loop and it must not long for long time, STP must detect the Loop and BLK port. 

[Joseph W. Doherty]

"Can Spanning Tree sometimes fail?"

My understanding is, it depends.

Depends on what?  Usually the portfast setting.

If portfast is active, STP skips the step of blocking the port until it's verified its activation will not create a L2 loop.  So, with portfast, a loop might be allowed.  However, STP is still active and will, if possible, detect the active loop and then block it.  The latter, though is not "guaranteed".  I.e. using portfast adds risk of a L2 loop being allowed.

[MHM Cisco World]

@Joseph W. Doherty right 
below is one caution form Cisco about portfast and it can cause temporary loop.

Caution Because the purpose of PortFast is to minimize the time that access ports must wait for STP to converge, it should only be used on access ports. If you enable PortFast on a port connected to a switch, you might create a temporary bridging loop.

for more info. 
check this link 
https://www.cisco.com/en/US/docs/general/Test/dwerblo/broken_guide/stp_enha.html

[IP_Cartel]

i've never seen it fail on networks ive been on.  i also set the data ports to spanning-tree portfast.

[abc1235]

@MHM Cisco World @Joseph W. Doherty and @IP_Cartel Thanks so much for the engagement. The temporary loop was exactly what happened in our situation, going by the following warning message when configuring portfast on an access port 

SW1(config-if-range)# spanning-tree portfast trunk
%Warning: portfast should only be enabled on ports connected to a single host. Connecting hubs, concentrators, switches, bridges, etc... to this interface when portfast is enabled, can cause temporary bridging loops. Use with CAUTION

So when the 2 switches were connected, a loop was created that was detected and corrected.

[Joseph W. Doherty]

As of "general rule of thumb", we don't enable port-fast on trunk ports.

Biggest reason we enable port-fast on access ports, is because user hosts tend to be powered on/off daily and w/o portfast DHCP might timeout and/or there's a noticeable delay to users before they can access network resources.

[abc1235]

Oh thanks for that @Joseph W. Doherty I just read more in detail about it. Thanks for that