Spanning Tree Protocol - Root Bridge Election

dipenshah · February 23, 2016, 11:41:36 AM

Spanning Tree Protocol - Root Bridge

The switch with the lowest MAC address value is elected as the root bridge (where priorities are equal)

Is there a reason why lowest MAC address value is always used? OR It's just a norm(rule) one should follow?

I feel it's more like a rule, but then every rule has some reasons. Can someone kindly explain or give a good reason why the protocol takes lowest MAC address as a root bridge?

Thanks

dlots · February 23, 2016, 11:49:26 AM

It takes the lowest bridge ID, that bridge ID is made of
1. Vlan
2. Priority
3. Mac-address

If you have ever wondered why the priority is basically in multiples of the max number of vlans this is why.

So The Vlans should always be the same, Priority is the same if you don't edit it, so that just leaves the MAC address as the tie breaker.

that1guy15 · February 23, 2016, 12:13:14 PM

I cant remember if I have had this explained or not. I think lowest was just chosen and we moved on. I also want to thing the oldest devices (lowest hardware address) was considered the stablest so thats why. I could be wrong all around.

But check out this presentation as it has a lot of good information about STP by its creator Radia Perlman
https://www.youtube.com/watch?v=L_zacX9DcZA

Might also try digging through some of the STP RFCs for an explanation.

routerdork · February 23, 2016, 01:42:34 PM

I would think it is to keep from having re-elections as newer devices were added. #termlimitsforrootbridges

icecream-guy · February 23, 2016, 02:18:17 PM

Quote from: that1guy15 on February 23, 2016, 12:13:14 PM
the oldest devices (lowest hardware address) was considered the stablest so thats why

That's what I was going to say.

dlots · February 23, 2016, 02:19:58 PM

Haven't gotten all the way though the video, but I really like it so far

NetworkGroover · February 23, 2016, 04:06:20 PM

What if a manufacturer has a lower OID? Would it really matter then how old a device was?

EDIT - Wow, OID? Really? I think I mean OUI? lol

srg · February 23, 2016, 04:12:41 PM

Quote from: that1guy15 on February 23, 2016, 12:13:14 PMMight also try digging through some of the STP RFCs for an explanation.

Actually STP is an IEEE standard, not IETF

. It should be available for download from ieee.org though.

srg · February 23, 2016, 04:13:48 PM

Quote from: AspiringNetworker on February 23, 2016, 04:06:20 PM
What if a manufacturer has a lower OID? Would it really matter then how old a device was?

EDIT - Wow, OID? Really? I think I mean OUI? lol

Lower OUI, you win.

NetworkGroover · February 23, 2016, 04:22:48 PM

Quote from: srg on February 23, 2016, 04:13:48 PM
Quote from: AspiringNetworker on February 23, 2016, 04:06:20 PM
What if a manufacturer has a lower OID? Would it really matter then how old a device was?

EDIT - Wow, OID? Really? I think I mean OUI? lol
Lower OUI, you win.

Exactly - so it's not about the age of the device.

I mean if I think about it longer than 2 seconds... I'd imagine they did what a lot of folks did back then.. just picked some shit and it worked.

I rarely listen to podcasts.. but listened to a very interesting one from PacketPushers talking about the future of networking with a gentleman who's been there since the beginning and talking about how IPv6 may *gasp* not really be needed for a few more decades... I don't remember his exact verbiage but it was something along the lines that they didn't really know what they were doing either so they just came up with something and if it worked they went with it.

I imagine there wasn't a whole lot you could use for elections that were static... burned-in MAC addresses seem like a good idea to me barring nothing else available... /shrug

EDIT - Did I just miss sarcasm there?

wintermute000 · February 23, 2016, 04:31:00 PM

Quote from: srg on February 23, 2016, 04:12:41 PM

Quote from: that1guy15 on February 23, 2016, 12:13:14 PMMight also try digging through some of the STP RFCs for an explanation.
Actually STP is an IEEE standard, not IETF . It should be available for download from ieee.org though.

What are you, the CWNP foundation??? (shakes fist at retarded exam questions)

srg · February 23, 2016, 04:41:21 PM

Quote from: wintermute000 on February 23, 2016, 04:31:00 PM
Quote from: srg on February 23, 2016, 04:12:41 PM

Quote from: that1guy15 on February 23, 2016, 12:13:14 PMMight also try digging through some of the STP RFCs for an explanation.
Actually STP is an IEEE standard, not IETF . It should be available for download from ieee.org though.

What are you, the CWNP foundation??? (shakes fist at retarded exam questions)

[emoji1] [emoji61]

RoDDy · February 25, 2016, 05:07:56 AM

Quote from: ristau5741 on February 23, 2016, 02:18:17 PM
Quote from: that1guy15 on February 23, 2016, 12:13:14 PM
the oldest devices (lowest hardware address) was considered the stablest so thats why

That's what I was going to say.

I believe this is correct. Oldest was chosen so that each time a new device was added, the network wouldnt need to be disrupted. The point about another manufacturer having a lower OUI is also valid which probably means they didnt think ppl would mix and match stuff. i dunno.

dlots · February 25, 2016, 08:02:49 AM

Honestly I don't think 50 sec of down-time was a big deal back then.

NetworkGroover · February 25, 2016, 09:44:32 AM

Quote from: dlots on February 25, 2016, 08:02:49 AM
Honestly I don't think 50 sec of down-time was a big deal back then.

Heh, truth - especially when the alternative was a broadcast storm or no redundancy.