HP procurve 1810-XXG trunking

M

morbid_looney

n00b
Joined
Feb 29, 2012
Messages
36
Hello there,


Recently I bought a HP procurve 1810-8G to go with my HP procurve 1810-24G.
The idea is simple the 1810-8G is on the ground level next to my router. and the 1810-24G is in the basement next to the servers.
I have 2 CAT6A cables connecting the switches.

First I tried it without changing any setting (you never know)
This created a loop and overloaded the entire network :)


so no luck on the easy approach, then I decided to trunk them.
I set up a trunk on both 1810's, and assigned the right ports.

But for some reason even though the ports where showing up as connected I could not get the trunk activated.

I had it set up as a dynamic trunk type, but since then I think I tried every single trunk combo.



I know that harforum recommends the 1810 series so I though id just ask here and see If someone could help me on this.


All questions are welcome.
I'm sure there you need a lot more info then what I just gave you, so just ask and I'll let you know.

PS:
There is no such thing as a obvious/ stupid question.
If you want to be sure I didn’t to something noobish just ask.


morbid_looney / koen
 
Last edited:
I am not too familiar with HP procurves, but trunking is trunking. Are you attempting to trunk with one connection, ie...Cat5 from one switch to the other? Or are you attempting to trunk via two Cat5 connections between the devices? Anytime you have more than one connection from one switch to another, STP will kick in and block one interface to prevent layer 2 loops. This is a good thing. If you have an unmanaged switch, well then you are SOL..haha

Start off with the simplest of configs. Connect one Cat5 to the switch and the other end to the other switch. I wouldnt use Dynamic trunking...setup the parameters on your own so you can tweek setting if need be. Each interface will need to be configured as a trunk and Dot1.q encapsulation turned on, if its not by default after commiting the ports as a trunk. Your trunk link should be up at this point, provided HP doesn't require some weird setting that I am not aware of. If you want another trunking interface between the two for redundancy and or bandwidth aggregation, then you need to logically combine the interfaces in use via LACP. This way here STP will not see these combined interfaces as a layer 2 loop. That will take a lot more configuration to get working. I hope I understood your question correctly and answered in a way that will help you out.
 
Last edited:
thank you for the reply, i will give it a shot right now.

PS the goal is bandwidth aggregation.
1GB/s isn't enough anymore.
 
And when I mean bandwidth (link) aggregation, that is meant soley between the two switches. ie..2 gig lag group between the switches. Are you running VLANs at all? If not, then simply connecting the two switches together via CAT5 should work as well if you are looking for port expansion. You dont necessarily need to configure these interfaces as "trunking" interfaces if you are not running multiple VLANs across those links.

When you say 1g is not enough.....do you mean you need more ports becasuse you ran out of ports on the switch, or do you mean you want more than 1g bandwidth between the two switches? Just wanted to clear that up....If its the latter, you can add more ports to the lag group than just two, provided you have the ports available.
 
Currently, i am not running any VLANS, but i am planing to do so in the future.
hence is why i choose for a trunk instead of a simple link aggr.

And yes i meant more bandwidth between the switches, the current bandwidth was insufficient..


Anyway, thank you very much for your help, the trunk is up and running now with 2 lines.

I started out with one like you said and then added the second.
In the end i don't see why it didn't work the first time as the setting seem to be exactly the same as i had it once before, only this time the status is active :)



thank you very much mikey for your great help.

As far is im concerned this threat is solved,



PS:
im not familiar with the rules of a solved threat.
do i have to put solved in the topic title or anything?
 
You don't have to put anything in the title of your thread to say "resolved." I would run some spanning tree commands on each switche to ensure that, those interfaces are in fact up. As previously stated, STP will "block" one interface if it detects a layer 2 loop which, truth be told, you just created. I know it appears that all the interfaces are up from a visible standpoint, but double check via some spanning tree commands. If you didn't logically combine those interfaces in a LAG group, i would be very suprised if all interfaces were up and functional. My guess is that one interface is in a blocking state.
 
Im a bit of a noob with networking (just a hobby) and i don't know how to run a spanning tree command :(

What i can tell you though is that all the ports a sending/receiving packets at the same rate on both switches, so data i flowing through both line.

But im not sure if that if that would rule out a block on one of the lines.

EDIT: MY PREVIOUS STATEMENT IS INCORRECT.

it seems that indeed one of the lines is sending/receiving a lot less packages
 
Issue a show spanning-tree command on both switches and check to see which interface on which switch is being blocked...Also, as with most vendor's CLI...type in either ? or a piece of the command_ ? and it should bring up the help menu and show you what other commands you can issue.
 
i am going to try it when i get home but im not sure if there is a CLI or console for the switch, in none of the documentation could i find anything about it.
ill look again when i get home.

EDIT:
it seems to support SNMP (read only) ill see what i can do with that.
 
Last edited:
You'll need to get on the CLI on each switch and add the ports to a trunk group. Then you'll trunk the group, not the ports. You're not going to see a bandwidth boost from this. The only benefit will be when there's multiple data streams going through the link. Two computers will be able to utilize a full Gb link between the switches simultaneously, but you wont see one application use both links at once.
 
after spending some time looking at SNMP, it seems the 2 lines are both active, one is up-link other is down-link, im happy with this.

no ports are completely blocked.
 
when you get ready to add vlans....just tag the same vlans on both switches in which you wish to pass through your connection.

Trunking is a cisco term. HP and several other vendors use other terms. Yes, it's the same thing essentially, but when talking to Procurve people out there....they may not know what you mean when you say trunking unless they've messed with Cisco equipment.
 
DlStreamnet said:
Wouldn't an etherchannel style situation be better?
Click to expand...

I'm not familiar with that (i don't study IT, just a hobby)

The goal is/was to have a connection (preferably greater then 1GB/s) between 2 switches.
And have to ability to use VLAN though both switches.

Fail-over/redundancy is not the goal here.
Just improved bandwidth and easy VLAN setup


If there is a better way to achieve my goals feel free too elaborate.
 
See my earlier post- You wont see a bandwidth improvement unless you're running multiple data streams. You'd have to move to 10G to increase bandwidth over a single data stream.
 
You must log in or register to reply here.
Back
Top