Link to IRF White Paper
HPN’s Intelligent Resilient Framework (IRF) is an advanced technology that allows one to aggregate 2 or more switches into a single switching and routing system also known as a “virtual switch”. IRF is available on many of the HPN A-Series switches.
IRF offers the following key advantages:
- Load balancing, aggregation and redundancy at all levels and of all functions: Link, L2 and L3 Forwarding, Management and Security.
- Simplified LAN design: by aggregating switches at the core and distribution layer, the combination of MSTP and VRRP, with all the complexity it implies in design, implementation and maintenance, is no longer required to achieve load balancing and resiliency.
About This Configuration
This configuration will go through the basic set up two switches to form an IRF virtual switch. We will also implement an additional level of high availablity by simply creating multiple aggregated links using standard link aggregation. Once the configuration has been implemented and the devices have formed an IRF connection, we will verify IRF operation with a few commands.
Equipment Used in This Configuration
- (2) A5800-24G Part # JC100A
- (2) 0.65m Direct Attach Cables (DAC) Part # JD095B
IRF Diagram
IRF Virtual Device Configuration Procedure List
for better planning. Because the configuration of some parameters takes effect after device reboot, you
are recommended to first configure parameters, power off the devices, connect devices physically,
power on the devices, and finally the devices will join in the IRF virtual device automatically. After an
IRF virtual device is formed, you can configure and manage the IRF virtual device by logging in to any
device in the IRF virtual device. The operations you make take effect on the master, and will be applied
to the member devices in the IRF virtual device.
Step 1: Login to the switch through the console port
Step 2: Ensure that both switches are running the same software version
<H3C>system view
[H3C]display version
Step 3: Reset the configuration of the switches.
<H3C>reset saved-configuration
<H3C>reboot
Step 4: Assign a unit number to each S5800. Switch 1 or 2. (Later you will see the unit number on the right side the switch on the front panel led)
On unit 1:
[H3C]irf member 1 renumber 1
Warning: Renumbering the switch number may result in configuration change or loss. Continue?[Y/N]:y
On unit 2:
[H3C]irf member 1 renumber 2
Warning: Renumbering the switch number may result in configuration change or loss. Continue?[Y/N]:y
Step 5: Save the configuration and reboot the switches
[H3C]quit
<H3C>save irf.cfg
<H3C>startup saved-configuration irf.cfg<H3C>reboot
Step 6: Setting priority on Master S5800.
On unit 1:
[H3C]irf member 1 priority 32
Step 7: Shutdown the 10 Gbps port that will form the IRF Group (on both switches)
On Unit 1:
[H3C]int TenGigabitEthernet 1/0/25
[H3C]int TenGigabitEthernet 1/0/25
[H3C-Ten-GigabitEthernet1/0/25]shutdown
[H3C]int TenGigabitEthernet 1/0/26
[H3C-Ten-GigabitEthernet1/0/26]shutdown
On Unit 2:[H3C]int TenGigabitEthernet 2/0/27
[H3C-Ten-GigabitEthernet2/0/27]shutdown
[H3C]int TenGigabitEthernet 2/0/28
[H3C-Ten-GigabitEthernet2/0/28]shutdown
Step 8: Assign the 10 Gbps port to an IRF port group
On Unit 1:
[H3C]irf-port 1/1
[H3C-irf-port]port group interface TenGigabitEthernet 1/0/25
[H3C-irf-port]port group interface TenGigabitEthernet 1/0/26
[H3C-irf-port]quit
On Unit 2:
[H3C]irf-port 2/2
[H3C-irf-port]port group interface TenGigabitEthernet 2/0/27
[H3C-irf-port]port group interface TenGigabitEthernet 2/0/28
[H3C-irf-port]quit
Step 9: Enable the 10 Gbps ports that will form the IRF (on both switches)
On unit 1:[H3C]int TenGigabitEthernet 1/0/25
[H3C-Ten-GigabitEthernet1/0/25]undo shutdown
[H3C]int TenGigabitEthernet 1/0/26
[H3C-Ten-GigabitEthernet1/0/26]undo shutdown
On unit 2:[H3C]int TenGigabitEthernet 2/0/27
[H3C-Ten-GigabitEthernet2/0/25]undo shutdown
[H3C]int TenGigabitEthernet 2/0/28
[H3C-Ten-GigabitEthernet2/0/26]undo shutdown
Step 10: Activate the IRF Port Configuration (on both switches)
[H3C]irf-port-configuration active
Step 11: Save the configuration
[H3C]quit
<H3C>save
Step 12: Connect the 2 10GbE Direct Attach Cables (DACs) as Shown in the IRF Diagram
NOTE: The secondary switch (unit 2) will now reboot automatically.
Step 13: The IRF stack should now be formed. Verify IRF operation
[H3C]display irf
[H3C]display irf configuration
[H3C]display irf topology[H3C]display devices
Step 14: Save the configuration
[H3C]quit
<H3C>save
Download the 58xx series IRF Configuration Guide for more information on configuring IRF including optional parameters.
***For all other switch models' IRF Configuration Guides and other manuals go to HP Networking main website. Enter "IRF Configuration Guide model number" in the search field top right corner of the page. For example: IRF Configuration Guide 5500EI***
Enjoy and thank you for your interest in IRF and HP Networking.
Where is the configuration file stored. Is it a single file as on a Cisco VSS or is there a configuration file kept per device ?
ReplyDeleteThe configuration file is maintained by the master IRF member and dynamically replicated to all other member switches in the IRF domain.
ReplyDeleteIf i need to delete device from XRN fabric and use default config. I need just disconnect it and use "reset Startup-configuration" command or some more manipulations with config files?
ReplyDeleteThanks.
Can implementing the performance too?
ReplyDeleteHow much stacking bandwidth with the IRF technology?
Up to 160Gbps
ReplyDeleteIf i have a third device....the port number will be ??
ReplyDeleteH3C]irf-port 3/3
[H3C-irf-port]port group interface TenGigabitEthernet 3/0/27
[H3C-irf-port]port group interface TenGigabitEthernet 3/0/28
[H3C-irf-port]quit
Regards,
Diogo
I just updated the IRF diagram to show a better representation of more than just 2 devices in an IRF group.
ReplyDeleteYou are correct with the TenGig interfaces starting with a 3 to represent the 3rd IRF member; however, the IRF port ould be 3/2 connecting to member 2's IRF port 2/1. Then to close the IRF ring you would connect 3/1 to 1/2. Again, see IRF diagram and assume there is no 4th IRF member in your configuration.
Hello Jason. Great write up. I have a few questions.
ReplyDelete1. Are the commands different when working with A7503s?
2 Do I need to be in chassis mode convert irf on a 7503?
Thanks.
Hello Anonymous, thanks for the feedback. Here's your answers...
ReplyDelete1. The only difference that comes to mind would be once 2 chassis are combined in an IRF group, there is now a chassis number added to the interface. For example, in an A5800-24G, the first gigabit interface of member 2 would be gig 2/0/1, the first number being the member ID, the second being subslot, and the third being the port. For an A7503, the first gigabit port of slot 1 on member 2 would be 2/1/0/1... 2 for the member ID, 1 for the slot, 0 for the subslot, and 1 for the port.... Answer 2 is also part of question 1.
2. "chassis convert mode irf"
Sets the Chassis to operate in IRF mode. Reboot the switch when prompted.
Additionally, make sure you have / purchase an A7503E for support of IRF. The A7503E-S does not support IRF.
Jason, great write up. One of our field engineers tried to use the guide on the H3C site and it is wrong, they googled and found yours and it worked great!
ReplyDeleteExcellent and concise guide. Cheers.
ReplyDeleteMore please!
Hello Jason,
ReplyDeleteHaving just sepnt a painful 48 hours following the configuration guides from H3C, we finally managed to get this working. I wish I had found your port explanation beforehand....would have made life a lot easier.
We have however found an issue stumping us,as the A7503E appears to have come up in a mode allowing it to see all the network traffic from anything even attached to one of the edge switches.
The Switch-mode, unchanged by our coming up in "Standard" and "Normal", but I cannot find a meaning for these terms the manuals.
This probably sounds very basaic, but it is my first attempt at this.
Have you any ideas how I could possibly get this 7503 acting as a switch? :)
Thanks for any ideas at all, or simply some direction for a newb....
Hi Pete,
Deletedid you resolve your problem?
I have two 4800-24G-SFP switches with the same problem.
I had around 400-600 packets/second arriving at my PCs interface (watching with Wireshark in promiscious mode) with traffic that definitley wasn't for my PC. Now it got better after enabling STP and setting STP parameters,
for some reason I was adviced to disable it during setup. Now I have only 10-20 packets/second.
I am 99% sure I have made some error in configuration, but have been over this, by the book so many times I'm probably not seeing the forrest for the trees!
ReplyDeleteI actually found a way in which the method described in the docs and this guide can lead to a failed cluster if your master switch faults out.
ReplyDeleteI also found a Best Practice to resolve that problem; a minor edit to your guide.
1) use a custom filename (e.g. hostname_irf.cfg) for the config on the master switch. use irf.cfg or similar for the rest
2) immediately after IRF reboots the slave switch:
- check the startup config setting on all switches.
- check the filesystems on each switch (e.g. dir slot2#flash:/)
3) now save the config on your master switch. If you are running 1211 or better it should prompt you to save on the other slots too. Now repeat above.
How I Came To Figure This Out:
At our site, on an 11XX firmware, after IRF was configured the old irf.cfg was left on the filesystem of switch2. It was actually booting from this and autoconfiguring via IRF every startup. So when switch1 failed, we lost the whole cluster. It actually flapped a bunch and came back up and then failed again a few days later. I was on vacation at the time so it was a sh*tstorm when I got back.
Do the check in #2 above on your system. You may be in the same situation I was and have only escaped the problem because your master node hasn't faulted (yet).
I was able to recreate this flaw consistently in the lab and my simplified BP above will avoid the problem.
Cheers.
P.S. Also I find it much easier to control things if you actually unpatch the IRF ports while doing this. Insert the patch cable/transceiver when you are ready to go. Otherwise you may find the slave switch will kick off after you have enabled only one of multiple IRF port group interfaces.
ReplyDeletedear Jason, im follow your tutorial but i cannot access the irf port
ReplyDelete[H3C]irf?
irf
[H3C]irf ?
auto-update Update application file automatically when versions mismatch
occurs
link-delay Link delay time
mac-address IRF MAC address
member Member of IRF
switch-to Switch to slave board
why i cannot go in command H3C]irf-port 3/3
please help
Just want to say thanks Jason for a easier guide to get started then from the guides I found.
ReplyDeleteBertahan,
I would make sure you use the command 'irf member 1 renumber 3' and save and reboot. Your switch should have all switch ports and irf port options set to 3/X/X or for IRF 3/3
Pretty well defined Guide from HP
http://bizsupport1.austin.hp.com/bc/docs/support/SupportManual/c02648772/c02648772.pdf
i am using HP A5500-24g EI swithc. is IRF is possible by doing virtual patch chord staking on dual personality ports.
ReplyDeleteOn unit 1:
ReplyDelete[H3C]irf member 1 renumber 1
By default, it's 1 - there is no need for this.
Jason & Group,
ReplyDeleteOne of my ten irf stacks just went into survivability, as switch 3 irf suddenly 'died'. Interestingly, the leds for the 4 sfp ports on the S5800-56C no longer light up (port state remains down) when physically plugged into any other switches ten gig ports. But when loop-backed, physically plugging in port 49 into port 50, the port states become active and the leds light up. Any ideas?
Robert Michael
RMR@pcsd1.net
Question regarding firmware upgrades of an IRF stack: Consider having 4 devices in the IRF stack. Now a new firmware has to be installed. There are documents about the upgrade, BUT: the IRF stack has to be rebootet! Does this mean a complete reboot, i.e. all devices at the same time?
ReplyDeleteThis breaks redundancy!
Or is it possible to do any kind of rolling upgrade? Unfortunately the docs are not describing, how to do a firmware upgrade without complete service disruption.
Any ideas? ...Thanks...
Hi Jason,
ReplyDeleteReally you have written a great blog for configuring the IRF. It really helps me a lot while doing the IRF cofig. Thanks once again for your great work...
Thank you Gaurav... I plan on adding more content by merging other information I have posted at various sites...
DeleteHi Jason,
ReplyDeleteA question on LACP-MAD. How frequently LACP-MAD packets will be advertised when there is IRF SPLIT?
Regards,
Kiran
Jason, I used your guide to connect together some H3C switches and some actual HP branded JC100A units that look exactly the same. However, I can't get the H3C and JC100A to play nice with each other. The H3C clearly has H3C firmware and the HPs have HP firmware. We tried loading the H3C firmeare on the HP unit with no luck. Any suggestions? I can't find anything that suggests these two shouldn't play nicely together.
ReplyDeleteYou will need to make sure both devices are running the same software version and both need to be running the HP version of the software. Both H3C branded and H3C branded switches can run the HP branded software, however, H3C branded software can only run on the H3C branded switches. Once you correct the software issue, both switches will be able to be in the same IRF group. Hope this helps.
DeleteJason, is it possible to add a description to the physical ports in an irf? something like
ReplyDeleteint ten1/2/0/1
desc IRF-PORT 1/1
The physical interface doesn't show up in disp conf.
Sorry, I meant disp current (HP Cisco blur). Answered my own question. The interface does show up but is pushed to the end of the config for some reason.
ReplyDeleteJason, I used your guide to connect 2 HP A5800-24G switches.
ReplyDeleteWorking fine. Problem is when I Power off the Unit-1, Unit-2 putting all its physical ports to shutdown mode,There is no service. If master fails whole stack is going down. Please give me any suggestion. Thanks in advance
Has anyone tried to setup an IRF over L2? Is it possible?
ReplyDeleteHi HP gurus, need help to find the cause and solution for the below error which I am getting when I am trying to setup stack through IRF on two A5800-JC100A.
ReplyDeleteI tried multiple reset saved-config and rebooted the device but getting the same error message if I try to give command "irf member 1 renumber 2 or 1". This is second unit and moreover when i run the command display irf this unit shows unit number2 and Master state. No IRF cable was connected.
I can able to configure unit 1 with the commands "irf member 1 renumber 1" "irf member 1 priority 10"
ERROR:- The device is not in the current IRF.
You may need to start over with the second switch. I would suggest you issue the "reset saved-configuration" command and to ensure you did not copy your running configuration to the defualt or null config.cfg file, go ahead and issue the "delete /un config.cfg" to permenetly delete the config.cfg file from flash. Issue the reboot command and the switch will generate a new and clean config.cfg file. Your switch will reboot in the default mode of IRF disabled and "member 1" Go ahead and follow the step above for configuring IRF.... Make sure connecting the cable is the last step you do after savng the config files etc... The second unit will reboot once connected to the IRF group.
DeleteI deleted the private-data and the saved-config and works, only saved-config doesnt work .
DeleteHi,
ReplyDeleteI am using HP A5500 EI 24G. I am planing to configure Active/Active IRF.
Please help me the configuration.
Better late than never... but hopefully you already had the chance to review the basics outlined in this post itself. Thank you.
DeleteI have an existing IRF deployment with 2x 5820. We're expanding and I'm debating whether we should add an additional 2x 5820 (if this is even possible without downtime) to this stack or setup another IRF on it's own. One of my main concerns is the fabric speed between the current and new switches (14 nodes @ 10G per switch in the current IRF stack).
ReplyDeleteYou have several options.. First, you can add the 2 new switches to the IRF group without an outage to the existing switches in the IRF group. Or, as you mention, you can create a new IRF group. The IRF links will not come into play unless there is an outage so long as you have dual links from your hosts, one to 2 different IRF members. Your choice on a larger IRF group vs. 2 IRF groups shuold be based on the type and flow of your data and the funtions of the 5820s (are they core, access TOR, etc.)
DeleteWe use these switches as all 3 roles. I guess the simpler question to ask would be would you recommend running iSCSI SAN traffic @ 10G across an IRF with 4 5820 members and 28 ESX hosts with 2 HBAs each?
DeleteThe iSCSI SAN is currently connected to 2 of the switch members and the HBA traffic does not currently traverse the IRF link that is established between switches.
Please correct me if I'm wrong here but I'm leaning towards an LACP trunk with at least 4 10G ports between 2 IRF groups to avoid the bottleneck the IRF link may create by leaving all 4 switches in a single group while keeping the SAN traffic in mind.
Your plan sounds good. Keep in mind, as you have already noted, in order to prevent traffic from traversing the IRF links (unless during an outage), proper design of uplinks use between IRF groups is important. It seems like you have this well under control. Good luck with everything and thanks for posting comments to this blog.
DeleteIs it possible to build IRF between 5800 and 7500? How is it different from Stacking? Can I connect FC SAN or FCOE SAN to 5800 switch? Do the switches support DCB?Can I configure the switches from Open Stack or some orchestration tool?
ReplyDeleteIRF groups can only be created within the same product model, e.g. 58xx models, 5500EI models, 5500HI models, etc. IRF between a 5800 and a 7500 is not supported.
DeleteThe 5900 now supports FCoE (Ethernet only) as well DCB protocols to allow for multihop. A future version of the 5900 may support native FC (under consideration at this time).
Please see the following whitepaper regarding HP's FCoE and data center convergence.
http://h20195.www2.hp.com/v2/GetDocument.aspx?cc=us&doclang=EN_US&docname=4AA4-5860ENW&doctype=white%20paper&lc=en&searchquery=&jumpid=reg_r1002_usen_c-001_title_r0001
Thanks,
Jason
Jason, can I made a IRF with same model but number of ports diferent ? like one 5120-48G-PoE and on 5120-24G-Poe ?
DeleteHi Jason, Hi Anonymus,
ReplyDeleteIs it possible to form an IRF stack with two
A7503-S chassis only by using the management module's interfaces? It comes with 24 SFP ports (17 to 24 dual personality ethernet ports). Is there any restrictions for using 10G interfaces only?
Hi Jason,
ReplyDeleteI am a newb on the H3C side. Have plenty of 3com/Comware switches, so the approach is familiar, but IRF is new for us.
Below is the config from the IRF ports, when I tried adding the IRF ports (backplane module JD360B) to the same IRF Port 1/1, the IRF fails and shuts down.
Should both ports off the same module be in the same IRF Port?
These are two A5500-HI switches.
interface Ten-GigabitEthernet1/1/1
#
interface Ten-GigabitEthernet1/1/2
#
interface Ten-GigabitEthernet2/1/1
#
interface Ten-GigabitEthernet2/1/2
#
irf-port 1/1
port group interface Ten-GigabitEthernet1/1/1 mode enhanced
#
irf-port 1/2
port group interface Ten-GigabitEthernet1/1/2 mode enhanced
#
irf-port 2/1
port group interface Ten-GigabitEthernet2/1/1 mode enhanced
#
irf-port 2/2
port group interface Ten-GigabitEthernet2/1/2 mode enhanced
#
return
Any advice? I think I am missing something...
Some switch models have specific IRF port requirements such as the 5500HI. It has "front side" and "back side" (via a module) 10G SFP+ ports.
DeleteThe following link is for the 5500HI IRF Configuration Guide. Skip down to page 4 under "Connection Requirements" section and you'll find what you are looking for.
"The SFP+ ports on the front panel and the
ports on the rear panel that are provided
by interface cards (if installed) can be
bound to any IRF ports." This refers to the "logical" IRF ports.
"The physical ports of an IRF port must
be located on the same interface
card.
• The SFP+ port numbered 53 on the
front panel and the physical ports on
interface card 2 must be bound to the
same IRF port.
• The SFP+ port numbered 54 on the
front panel and the ports on interface
card 1 must be bound to the same IRF
port." This refers to the "physical" IRF ports.
http://bizsupport2.austin.hp.com/bc/docs/support/SupportManual/c03180062/c03180062.pdf
Thanks for being an HP customer!
Jason
Jason, Thanks for the quick reply. Did did review that document. While it did not say that the ports had to be in the same group, it did not say they could be in separate groups either. I decided to be brave and attempt to add the ports to the SAME IRF port - config below.
ReplyDelete#
interface Ten-GigabitEthernet1/1/1
#
interface Ten-GigabitEthernet1/1/2
#
interface Ten-GigabitEthernet2/1/1
#
interface Ten-GigabitEthernet2/1/2
#
irf-port 1/1
port group interface Ten-GigabitEthernet1/1/1 mode enhanced
port group interface Ten-GigabitEthernet1/1/2 mode enhanced
#
irf-port 2/2
port group interface Ten-GigabitEthernet2/1/1 mode enhanced
port group interface Ten-GigabitEthernet2/1/2 mode enhanced
#
return
Now, the display IRF Topology command yeilds a somewhat concerning output.
[AMYO-SWSR1-IRF]dis irf top
Topology Info
-------------------------------------------------------------------------
IRF-Port1 IRF-Port2
Switch Link neighbor Link neighbor Belong To
2 DIS -- UP 1 d07e-286f-74ab
1 UP 2 DIS -- d07e-286f-74ab
Previously, when each physical port was contained within their (own) respective IRF Port, I would get Link UP accross the board...
So much for going production this weekend...
How many switches are you connecting in the IRF group? If only two then this is normal. Rememeber, there is a logical IRF port and physical 10G interfaces assigned to those logical IRF ports. The only reason you would see IRF port 2 on the first switch "up" is if you were creating an IRF "ring" from the 3rd or 4th physical switch (see diagram above). With only two switches, there is no need to create a ring, this is accomplished by having more than 1 physical port assigned to a logical IRF port (like link aggregation).
DeleteJason,
ReplyDelete2 switches. That makes sense. So effectively, I have created a ring if I use:
irf-port 1/1
port group interface Ten-GigabitEthernet1/1/1 mode enhanced
#
irf-port 1/2
port group interface Ten-GigabitEthernet1/1/2 mode enhanced
#
irf-port 2/1
port group interface Ten-GigabitEthernet2/1/1 mode enhanced
#
irf-port 2/2
port group interface Ten-GigabitEthernet2/1/2 mode enhanced
#
return
but, that is not neccessary and i can use:
irf-port 1/1
port group interface Ten-GigabitEthernet1/1/1 mode enhanced
port group interface Ten-GigabitEthernet1/1/2 mode enhanced
#
irf-port 2/2
port group interface Ten-GigabitEthernet2/1/1 mode enhanced
port group interface Ten-GigabitEthernet2/1/2 mode enhanced
And I can ignore the 'disconnected' links in each IRF port.
You got it!
Deleteirf-port 1/2 (switch member 1) and irf-port 2/1 (switch member 2) are not used and will show as "dis" status. If you add another switch then you will use these IRF ports as shown in the config example above.
Thanks for posting comments to the blog... Enjoy your new switches and good luck with everything else going prod this weekend.
Jason,
ReplyDeleteThanks! Would you be opposed to a quick review of my config? Just for any major gotchas that I may be missing? I don't have a consultant on this project and I am on my own....
Hi there, Just my 2c after following this guide for the past day. This works fine if you have two switches and two 10Gb modules. :)
ReplyDeleteIf you have 4 switches and still only one module per switch then assigning multiple virtual ports to an interface caused issues (for me anyway).
The way around it was to bind only 1 physical port to each virtual irf port (i.e. irf-port x/1 is bound to x/x/1, irf-port x/2 is bound to x/x/2) and then cable it up with:
1/1 going to 2/2
1/2 going to 4/1
2/1 going to 3/2
3/1 going to 4/2
Guide is here: http://cciestudy-cciejourney.blogspot.com.au/2011/12/stacking-on-h3c-s5500.html
This is long winded so grab a snack first :-)
ReplyDeleteI am setting up 3 A5500 switches to be deployed and I get duplicate IP address errors anytime I connect a device (PC) to one of the switches (actually it is 2 switches joined together by irf-port).
Switch1 132.156.60.3 /23
Switch2 132.156.60.13 /23
switch3 132.156.60.14 /23
Switch1 connected by Gi1/0/52 to switch2 Gi1/0/25
PC1 (connected to switch1 Gi1/0/40) 132.156.60.200 /23
PC2 (connected to switch2 Gi1/0/12) 132.156.60.100 /23
It only happens on switch1, I can clear the ARP table on all the switches and it does not make any difference. For some reason it always comes back saying that the address is already used on another port (the port is Gi1/0/52 and is the access link to switch2)
All switch ports on same VLAN, and no trunk ports.
There is nothing else connected on these switches and there are definitelly no duplicate addresses that would cause these errors.
I have tried with 2 different PCs that have different addresses and get the same result for both. Firmware was updated to the most current one an all the switches.
display version
HP Comware Platform Software
Comware Software, Version 5.20.99, Release 2220P02
Copyright (c) 2010-2013 Hewlett-Packard Development Company, L.P.
HP A5500-48G-PoE+ EI Switch with 2 Interface Slots uptime is 0 week, 0 day, 1 hour, 46 minutes
HP A5500-48G-PoE+ EI Switch with 2 Interface Slots with 1 Processor
256M bytes SDRAM
32768K bytes Flash Memory
Hardware Version is REV.C
CPLD Version is 002
Bootrom Version is 707
[SubSlot 0] 48GE+4SFP+POE Plus Hardware Version is REV.C
[SubSlot 1] 2 SFP+ Hardware Version is REV.A
[SubSlot 2] 2 SFP+ Hardware Version is REV.A
display arp
Type: S-Static D-Dynamic
IP Address MAC Address VLAN ID Interface Aging Type
132.156.60.200 0015-609c-c17c 60 GE1/0/40 18 D
132.156.60.13 d07e-2884-0c7f 60 GE1/0/52 4 D
132.156.60.100 001b-38be-da74 60 GE1/0/52 20 D
From console on switch1, I then ping directly connetcted PC1 ans immediatelly start to get duplicate IP errors
ping 132.156.60.200
PING 132.156.60.200: 56 data bytes, press CTRL_C to break
%Apr 27 10:13:19:380 2000 BCC10-RM150 ARP/5/ARP_DUPLICATE_IPADDR_DETECT: Detected an IP address conflict. The device with MAC address 0015-609c-c17c connected to GigabitEthernet1/0/40 in VLAN 60 and the device with MAC address 0015-609c-c17c connected to GigabitEthernet1/0/52 in VLAN 60 are using the same IP address 132.156.60.200.
Request time out
Request time out
Request time out
Request time out
Request time out
--- 132.156.60.200 ping statistics ---
5 packet(s) transmitted
0 packet(s) received
100.00% packet loss
display arp
Type: S-Static D-Dynamic
IP Address MAC Address VLAN ID Interface Aging Type
132.156.60.13 d07e-2884-0c7f 60 GE1/0/52 3 D
132.156.60.100 001b-38be-da74 60 GE1/0/52 20 D
132.156.60.200 0015-609c-c17c 60 GE1/0/52 20 D
If I disconnect Switch1 Gi1/0/52, no errors are reported. Any insights, help?
Luc,
ReplyDeleteUnfortunately I can only give you some helpful suggestions related to your IRF configuration. It's always best to engage with support just in case you have an issue that needs further troubleshooting. You are entitled to support with your lifetime warranty.
Briefly, it looks like your IRF configuration may not be correct. With the 5500EI you can only establish an IRF connection via 10G interfaces. You mention that you are connecting via 1G interfaces. As well, when IRF is established the switches will have different prepended numbers to represent the member ID. For example, switch 1 will be Ten-Gigabit1/1/1 and switch 2 will be Ten-Gigabit2/1/1 (with a 10G module installed in the back of both switches)
Please follow the directions included above for proper set up and configuration of IRF. Thanks Luc.
This comment has been removed by the author.
ReplyDeleteHi, Our customer have purchased the SFP cables for server connectivity and additionally I have only one cable with me and I am planning to configure the IRF in between 2 A5820 Switches, any possible to work?
ReplyDeletebelow are configuration I planned to do ..
[H3C]irf-port 1/2
[H3C-irf-port2/1]port group interface ten1/0/27
[H3C]irf-port 2/1
[H3C-irf-port2/1]port group interface ten2/0/27
Apologizing if it is stupid quest, because I am server engineer and just supporting as it is for best practice.
Nichu,
ReplyDeleteit shall work fine as you have mentioned...only challenge shall be a single point of IRF link failure which is not a recommended design...you shall configure IRF MAD to ensure that IRF link failure does not lead to multiple identical active IRF virtual devices...
Mahesh
Mahesh,
DeleteThanks for your response. I configured with single link which was working and requested to purchase additional cable to fulfill the requirement. Below are the conf details I made on switches.
irf member 1 priority 10
irf-port 1/1
port group interface Ten-GigabitEthernet1/0/1 mode enhanced
#
irf-port 2/2
port group interface Ten-GigabitEthernet2/0/2 mode enhanced
About IRF MAD : I have only two switches with me and I set the priority 10 for switch 1. Is this required to configured MAD in between this..?
I looking over some docs related to this, came to know that configuring LACP MAD - need 3rd switch and for BFD MAD - need separate link between switches (Lack of cable (possible to use of Gig port?)) and finally ARP MAD not widely deployed. Pls help me, how to configure MAD if necessary to implement in our system..
Your prompt response will be highly appreciated.
Nichu, did you conclude on your IRF MAD issue? I am currently faced with the same scenario? Cheers.
DeleteHi, I am facing an another error while doing the link aggregation.. below are brief..
ReplyDelete--------------------------------------------------
display link-aggregation verbose b3
Loadsharing Type: Shar -- Loadsharing, NonS -- Non-Loadsharing
Port Status: S -- Selected, U -- Unselected
Flags: A -- LACP_Activity, B -- LACP_Timeout, C -- Aggregation,
D -- Synchronization, E -- Collecting, F -- Distributing,
G -- Defaulted, H -- Expired
Aggregation Interface: Bridge-Aggregation3
Aggregation Mode: Dynamic
Loadsharing Type: Shar
System ID: 0x8000, 20fd-f1e4-5a01
Local:
Port Status Priority Oper-Key Flag
--------------------------------------------------------------------------------
XGE1/0/24 U 32768 2 {ACG}
XGE2/0/24 S 32768 2 {ACDEFG}
Remote:
Actor Partner Priority Oper-Key SystemID Flag
--------------------------------------------------------------------------------
XGE1/0/24 0 32768 0 0x8000, 0000-0000-0000 {EF}
XGE2/0/24 0 32768 0 0x8000, 0000-0000-0000 {DEF}
%Apr 30 13:57:19:023 2000 A5820X-IRF IFNET/3/LINK_UPDOWN: Ten-GigabitEthernet1/0/24 link status is UP.
%Apr 30 13:57:19:024 2000 A5820X-IRF LAGG/5/LAGG_ACTIVE: Member port Ten-GigabitEthernet1/0/24 of aggregation group BAGG3 becomes ACTIVE.
%Apr 30 13:57:19:024 2000 A5820X-IRF LAGG/5/LAGG_INACTIVE_PARTNER: Member port Ten-GigabitEthernet2/0/24 of aggregation group BAGG3 becomes INACTIVE because the configuration of its partner is improper for aggregation.
%Apr 30 13:57:19:219 2000 A5820X-IRF LLDP/6/LLDP_CREATE_NEIGHBOUR: -Slot=1; New neighbor created on Port Ten-GigabitEthernet1/0/24 (IfIndex 11796503), Chassis ID is e41f-13d8-2626, Port ID is e41f-13d8-2626.display link-aggregation verbose b3
%Apr 30 13:57:20:363 2000 A5820X-IRF OPTMOD/4/MODULE_IN: -Slot=1;
Ten-GigabitEthernet1/0/24: The transceiver is STACK_SFP_PLUS.
%Apr 30 13:57:20:771 2000 A5820X-IRF IFNET/3/LINK_UPDOWN: Ten-GigabitEthernet1/0/24 link status is DOWN.
%Apr 30 13:57:20:772 2000 A5820X-IRF LAGG/5/LAGG_INACTIVE_CONFIGURATION: Member port Ten-GigabitEthernet1/0/24 of aggregation group BAGG3 becomes INACTIVE because the port's configuration is improper for aggregation.
%Apr 30 13:57:20:772 2000 A5820X-IRF LAGG/5/LAGG_ACTIVE: Member port Ten-GigabitEthernet2/0/24 of aggregation group BAGG3 becomes ACTIVE.
%Apr 30 13:57:21:075 2000 A5820X-IRF IFNET/3/LINK_UPDOWN: Ten-GigabitEthernet1/0/24 link status is UP.
%Apr 30 13:57:21:076 2000 A5820X-IRF LAGG/5/LAGG_ACTIVE: Member port Ten-GigabitEthernet1/0/24 of aggregation group BAGG3 becomes ACTIVE.
%Apr 30 13:57:21:076 2000 A5820X-IRF LAGG/5/LAGG_INACTIVE_PARTNER: Member port Ten-GigabitEthernet2/0/24 of aggregation group BAGG3 becomes INACTIVE because the configuration of its partner is improper for aggregation.
%Apr 30 13:57:21:222 2000 A5820X-IRF LLDP/6/LLDP_CREATE_NEIGHBOUR: -Slot=1; New neighbor created on Port Ten-GigabitEthernet1/0/24 (IfIndex 11796503), Chassis ID is e41f-13d8-2626, Port ID is e41f-13d8-2626.
--------------------------------------------------
I have connected with server. individually both the links are working. When connecting to Aggregated ports one is automatically moving to Non-selected mode.
What might me the reason, what are the steps I can take to overcome this ?
Nice Blog ,Thanks for Sharing the information.Customer Satisfaction is very important to grow up the business.
ReplyDeletei am tring to configure A5500 series witch for irf.. i followed your configuraion steps but finaly i connect the cable it says " IRF port 1 is down because heartbeat timed out." why is that? i am using SFP+ for link two switches
ReplyDeletehi
Deleteconfigure ports as below your issue will be resolved
in master switch physical port 53 = irf port 1/1
in slave switch physical port 54 = irf port 2/2
do the cross connection with DAC cable .
Farzan
Hello, you could support a Port-security tutorial. I have an HP A5120 EI.
ReplyDeleteMy email is adonaiferia@gmail.com
The main thing is:
Allow only one MAC address on a port
Disable the port if it detects two MAC Address
How to detect a blocked port
How to enable port
Thanks and hope you can help
thank you.
ReplyDeleteHi,
ReplyDeleteGreat write up.. can you please guide me in the steps if there is one member switch in the irf stack goes faulty. Then how do I put a new one withuout disrupting the network or getting the active switch rebooted..
Appreciate your prompt response...
Hi i am having two A5500-EI 48G port switches and i am trying to do IRF configuration by using DAC cable (SFP+)
ReplyDeletewhen i want to assign physical port to IRF port i can only add one physical port. when i going to add another port it says
ERROR: This port can't be added to this IRF port group!
why ?????
check it port is unshut.
DeleteIt has to be shutdown while binding to IRF Port
I'm thinking to build an IRF domain with 3 pcs. HP 5500-24G-4SFP.
ReplyDeleteBut one thing which I don't completely understand is what happens if master switch fails and how long does it take to the rest of the switches to return to normal operation.
It is said at IRF configuration guide that: "After a master election, all slave member switches initialize and reboot with the configuration on the master, and their original configuration, even if has been saved, will be lost."
Does this mean that one of the rest 2 switches that would take the master role would continue operating normally but the 3rd switch would have to reboot? Or the 3rd switch could be assigned to new master without a reboot as it will have the same configuration?
hi mr. works with 3com switch 4500 series?
ReplyDeleteHi,
ReplyDeleteI have a live standalone HP5800, my network has expanded and I have purchased an additional 5800 switch. Is it possible to configure IRF without downtime on the first switch?
You will need to reconfigure both switches to work in "IRF" mode, create the IRF group, etc. This will require a reboot of both switches.
DeleteYou will need to reconfigure both switches to work in "IRF" mode, create the IRF group, etc. This will require a reboot of both switches.
ReplyDeleteis their a way to send port error/ port down messages to a management overseeing program?
ReplyDeleteHi Jason,
ReplyDeleteCan we create IRF between HP 5900 HP 5920 switches? Please help
Yes, you can, but you need to configure 4 Ports as IRF. Even when you only using 2 Ports for the link.
DeleteMy procure switch was acting strange...the port lights were on without any ethernet cables plugged in. Anyone know about this issue and how to resolve.
ReplyDeleteHi, when we connect our second switch using the IRF cable it is causing a power surge which is killing our server and devices. This occurs with switches grounded and ungrounded. Any suggestions?
ReplyDeleteHi Jason,
ReplyDeleteCan we create IRF over Fiber also ? We have fiber with length more than 200 yards and want to implement IRF with having HP core on both sites
Hello, We have an IRF between 3 switches (HP 5130) as our access layer. 2 of them are 48ports and one is 24ports. This has always worked. Until the power was disconnected on one of the switches. After a reboot of the stack the 24p switch did not join the IRF stack. We checked the config with a console cable and the switch was renumbered from 3 to 1. AND it was trying to join the stack on 1/0/50 and 1/0/51. witch did not work because it has only 24 ports... We fixed it by reconfiguring.. It looks like it took the config from the master 48p switch and flipped. For this reason is it not recommanded to use switches with differen port numbers?
ReplyDeleteThank you! nice article :)
thanks a lot
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteIf I use FortyGig cabling on 5900AT (4 switches stack), can I then extend it to additional switches in DIFFERENT building in any way?
ReplyDelete
ReplyDeleteThank u for the wonderful post !!
Hp Printer Technical Support
This is an excellent writeup Jason and I hope to follow it soon. I have three 5130's that I plan to ring in IRF, however I was also planning on connecting our two VM hosts via 10G to our "top of rack" 5130 Master, but the IRF will eat up all four SFP+ ports on all switches. Without two free SFP+ I will have to go back to LACP on four gig Ethernet (8 ports gone for this). If I wanted to add another switch that will give me more 10G links, how homogeneous does it have to be to the 5130? What are the potential risks?
ReplyDelete-Eli
I have two HP5940 switches, after IRF configuration its showing " The port Ten1/0/48 can't receive irf pkt, please check."
ReplyDeletePlease help
Make sure Irf configuration active command ran on the switch...
DeleteHi, thanks for this article which helps a lot to understand. I have just one question, i have to add one nex switch to a stack of IRF.Do i just cable it and it's ok or do i made some other configuration?
ReplyDeleteThanks a lot
Hi,
ReplyDeleteFirst thanks for this guide!
I have two 5940 48sfp+ 6qsfp28. I doing IRF with sfp+ port and no problem.
But when I want doing with qsfp28 port (40gb cable DAC), impossible up link of IRF.
I update with last firmware (7.1....r2509P02), but problem persist.
Anyone can help me please?
Thanks!
I have two 5940s with with an IRF that appears to be fully functional. I have two uplinks, and the remote side is managed by VRRP, and I use their VIP as my default route. For some reason, a seemingly random set of IPs cannot reach the other side of the active uplink - but all IPs can see our end of the uplink and vice versa. Taking either uplink down changes the behavior. Any ideas on what could contribute to that behavior ?
ReplyDeletei have 3 switches connected alrady by stacking cable not spf port and i need to add new switch but not working after compelte all steps
ReplyDelete1- upgrade as the same switches
2 get it member 4 and priority 26 note master:32
3 irf domain as the same
4 reset saved configuration
i need what,s the problem about configuration
sory model of device : hp A5500 comware
ReplyDelete
ReplyDeleteHi there, awesome site. I thought the topics you posted on were very interesting. I tried to add your RSS to my feed reader and it a few. take a look at it, hopefully I can add you and follow.
HP Officejet 3830 Wireless Setup
ReplyDeleteI know Laura well and she is principled, thoughtful, and extremely bright...more power to her!
123 HP Setup
Really its very useful information that you have shared and thanks for sharing the information with us.
ReplyDelete123 HP Deskjet Setup 5820
Thank you for the insights offered in the article. Looking for more such pieces. Visit our sites for printer setup & support.
ReplyDelete123.hp.com || 123.hp.com/setup || 123HP Setup || hp.com/setup || hp.com/123 || 123.hp.com setup || 123 HP Printer Setup || 123 HP Printer Support || 123 HP Setup and Install || 123hpcom || 123 HP Printer Install || 123hpcomsetup || 123 HP Wireless Setup || 123 HP Install || hpcom/123 || 123hpcominstall || 123HP Setup || 123 HP Smart App || Install 123 HP Printer || HP 123 Setup Scanner
Ohh, great. Thank you for sharing.
ReplyDeleteAntivirus || Antivirus Software || Antivirus Cleaner || Internet Security || Antivirus Protection || Internet Security Software || Antivirus Download || Security Antivirus || Antivirus and Security Software || Install Antivirus || Antivirus & Security Software || Mobile Security and Antivirus ||
123 hp com setup
ReplyDeleteDIGITAL MARKETING COMPANY MEERUT
Lexmark Printer Support
www office com setup and follow the on screen instructions
HP A5120 JG237A IRF CONFIG ISSUES. Replacing a member switch in an irf stack of 3 switches but having issues with the master picking up the slave.
ReplyDeleteHP A5120 JG237A IRF CONFIG ISSUES.
Replacing a member switch in an irf stack of 3 switches but having issues with the master picking up the slave.
Started new job and have never touched HP A series switches before....
Process completed:
Added same firmware as master.
The replacement member is 3, so renumbered to 3
Changed priority to match being 3rd in the stack ie 30 - save, reboot.
Cofigured irf
[switch3] int ten 3/2/1
[switch3] shutdown
[switch3] int ten 3/2/2
[switch3] shutdown
[switch3] irf-port 3/1
[switch3] port group int ten 3/2/1
[switch3] port group int ten 3/2/2
[switch3] int ten 3/2/1
[switch3] undo shutdown
[switch3] int ten 3/2/2
[switch3] undo shutdown
[switch3] irf-port-configuration active
[switch3] quit
[switch3] save
powered off the switch and put in the irf cables. I know that the irf cables are in the correct place as we have good documentation and have tried swapping around, but when I boot the new switch it does not pick up the config from the master.
Currently the irf stack looks like below, but irf has been configured on the new switch and cables are plugged in.
It's as though it just isn't picking up the irf connections. It really doesn't make any sense and would like to know if I am missing something here?
1 1 Down Ten-GigabitEthernet1/2/1
1 2 Up Ten-GigabitEthernet1/2/2
2 1 Up Ten-GigabitEthernet2/2/1
2 2 Down Ten-GigabitEthernet2/2/2
Thanks and that i have a swell give: How To Plan House Renovation exterior home renovation
ReplyDeleteHello everyone,
ReplyDeleteI hope my message will be read.
In my IRF, I have 7 members
As part of a move within the same datacenter, I had to move members 1, 2, 3, 4 but before switching them off, I changed the priorities :
Initial priorities
irf member 1 priority 32
irf member 2 priority 30
irf member 3 priority 28
irf member 4 priority 26
irf member 5 priority 24
irf member 6 priority 22
irf member 7 priority 20
Modified priorities
irf member 1 priority 26
irf member 2 priority 24
irf member 3 priority 22
irf member 4 priority 20
irf member 5 priority 28
irf member 6 priority 30
irf member 7 priority 32
I turned off member 1, 2, 3 and 4, then moved them to the new location.
I racked member 1 and connected :
Member 1 - 1/0/21 to Member 7 - 7/0/23
Member 1 - 1/0/22 to Member 7 - 7/0/24
I started up member 1: Result OK, after a few minutes it synchronized.
I racked member 2 and connected :
Member 2 - 2/0/21 to Member 1 - 1/0/23
Member 2 - 2/0/21 to Member 1 - 1/0/24
I started up member 2: Result OK, after a few minutes it synchronized.
I racked member 3 and connected :
Member 3 - 3/0/21 to Member 2 - 2/0/23
Member 3 - 3/0/22 to Member 2 - 2/0/24
I started up member 3: Result OK, after a few minutes it synchronized.
I racked member 4 and connected :
Member 4 - 4/0/21 to Member 3 - 3/0/23
Member 4 - 4/0/22 to Member 3 - 3/0/23
I started up member 4: Result OK, after a few minutes, it synchronized.
But when I connected member 4 and member 5 (the ports were active), the IRF malfunctioned. The members all rebooted.
Member 4 - 4/0/23 to Member 5 - 5/0/21
Member 4 - 4/0/24 to Member 5 - 5/0/22
So I disconnected Member 4 from connections 4/0/23 and 4/0/24, then switched off Members 1, 2, 3 and 4.
Can you please tell me what happened.
Thank you in advance and I look forward to hearing from you.