Brocade ICX Series (cheap & powerful 10gbE/40gbE switching)

[dbvader]

Well, now I tried my luck with one of my old computers having a physical RS-232 interface on its motherboard and guess what, it prints garbage to the console (I've triple checked the connection settings).

Feels like it's the switch and not the cable or host. Can you perform a factory reset?

 

[rnavarro]

Friends don't let friends do IPv6

Ha! IPv6 is the beez neez

No NAT, direct connectivity, TONS of IPs that I can route with at will, SLAAC is great.

I'm sold on it

I have limited IPv4 space at my colo so I was forced to learn more about IPv6 and I'm definitely onboard now

 

[fohdeesha]

Well, now I tried my luck with one of my old computers having a physical RS-232 interface on its motherboard and guess what, it prints garbage to the console (I've triple checked the connection settings).

try these baud rates instead, one at a time: 14400, 19200, 38400, 57600, 115200

on the uboot/linux based models like the 6450, it's possible to change the serial baud rate in the u-boot environment variables, I suppose it's possible someone did this previously. I have only ran into it once though, previous user forced it to 115200

 

[Testsubg]

Load address: 0x3000000
Download to address: 0x3000000
Loading: %T %T %T %T %T %
Retry count exceeded, Try again
ICX64XX-boot>>
[/CODE]

I’ve connected to the TFTP server fine and update_primary without issue, but the uboot update always fails. Was this ever resolved/or how did anyone else fix it?
On mine it says:

Download Filename 'ICX64xx/kxz10105.bin'
Load address: 0x3000000
Download to address: 0x3000000
Loading: %T %T %T %T %T %
Retry count exceeded, Try again
ICX64XX-boot>>

Also says,

update_uboot
egiga0 no link
Using egiga0 device

Seems like there’s something hopefully simple that I’ve missed. I appreciate any help with this!

 

[egyptianthemagician]

Looking for some help. I've been using the brocade6450 as a basic network switch for the past year. Didn't touch any settings at all. Recently upgraded to 3gb internet so wanted to use the 10G SFP slots. I bought two of these from Amazon - https://a.co/d/aFhYnIa

Have this connected to my unraid server which shows a 10000 Mbps eth0 link (previous was 1000 Mbps) so I thought everything was groovy. When I did a speed test though my download and upload fluxuate greatly.




This is using a docker tool called speed test tracker does speed tests every 30 mins

My download goes up and down hitting close to max of 2500 (since the router is a 2.5 GB router) but my upload seems to cap below 300.

Wondering if anyone had experience with these trancievers or any thought on what's going on.

 

[Catlike]

Looking for some help. I've been using the brocade6450 as a basic network switch for the past year. Didn't touch any settings at all. Recently upgraded to 3gb internet so wanted to use the 10G SFP slots. I bought two of these from Amazon - https://a.co/d/aFhYnIa

Have this connected to my unraid server which shows a 10000 Mbps eth0 link (previous was 1000 Mbps) so I thought everything was groovy. When I did a speed test though my download and upload fluxuate greatly.


View attachment 33101

This is using a docker tool called speed test tracker does speed tests every 30 mins

My download goes up and down hitting close to max of 2500 (since the router is a 2.5 GB router) but my upload seems to cap below 300.

Wondering if anyone had experience with these trancievers or any thought on what's going on.

Are you able to run a Speedtest directly on your router to set a baseline before moving onto clients?

that is where I would start.

 

[egyptianthemagician]

Are you able to run a Speedtest directly on your router to set a baseline before moving onto clients?

that is where I would start.

Upload speed is not great directly on the router so will probably have to dig into that as well.

 

[bwahaha]

I’ve connected to the TFTP server fine and update_primary without issue, but the uboot update always fails. Was this ever resolved/or how did anyone else fix it?
On mine it says:


Also says,
Seems like there’s something hopefully simple that I’ve missed. I appreciate any help with this!

Shot in the dark, but I think egiga0 is the Management ethernet port (might be wrong). The error says it's not plugged in. Is that the port you've plugged into the tftp server? Cable good?

 

[ManoftheSea]

Hey network people.
I've been having some adventures with an ICX6450 and a Ubiquiti AP and a Roku. An old roku stick, if it matters. I had it set up that the ICX6450 has DHCP pools, the AP is connected on a dual-mode port, with untagged traffic for the AP itself and tagged VLANs for the wifi. Alas, I think it was stated a bit ago, and I discovered as well, that the Roku is unwilling to get an address from the ICX6450.

So I set up a DHCP server (Kea) on a host, and I configured dhcp snooping ("ip dhcp snooping vlan N"), while also enabling dhcp trust ("interface eth 1/2/4", "dhcp snooping trust"). And when I set up the helper address ("ip helper-address 1 addr.of.host"), I see WIRED devices able to get IPs from Kea. But... I don't see wireless clients getting addresses (and therefore not finishing the association process). tcpdump from the Kea server shows the DISCOVER and OFFER; tcpdump from a wireless client even showed the REQUEST being sent, but it doesn't arrive.

Does anyone have a suggestion? Maybe something at the ubiquiti AP configuration is causing DHCP to the VLAN's VE to work, but not to get relayed? I'm sure this is a scenario that plenty of people have working, what am I missing?

So it turns out, if I have a DHCP-server pool configured for the address space, then the dhcp relay will not work. I had thought that putting the pool in "pending" status ("no deploy") would be enough. But when I undefined it ("no ip dhcp-server pool guests"), the relay immedately started working.

Friends don't let friends do IPv6

Oof, man. How are we ever going to deploy a 20 year old technology with that attitude?

 

[1vivy]

Well I upgraded to 09.x branch and I'm still seeing the same issues

I guess i'll just switch to moving the L3 terminations off the switch as it's super annoying to go up and down stairs losing my IPv6 connection (which is favored)

If it's okay, could you share your L3 config on the switch? Do you have a static IPv6 prefix? I'd like to experiment with putting my IPv6 routing on the switch, currently just have IPv4 on the switch and VyOS handling IPv6 routing.

 

[rnavarro]

If it's okay, could you share your L3 config on the switch? Do you have a static IPv6 prefix? I'd like to experiment with putting my IPv6 routing on the switch, currently just have IPv4 on the switch and VyOS handling IPv6 routing.

My IPv6 prefix was "static" - technically the ISP could renumber me, but they never have.

Here is the config I used prior to removing all the L3 stuff.

ICX 7250 Config

GitHub Gist: instantly share code, notes, and snippets.

gist.github.com

I used an upstream OPNSense router to do my WAN routing/firewalling. This was piped down with a transit vlan (vlan 88 in the config)

I don't think there was anything crazy in my config that was preventing the IPv6 neighbor cache from automagically updating, but who knows.

I've already given up on it and I ended up hosting a pair of VyOS instances as my "core" router and doing the same upstream transit to my OPNsense router.

Edit - and yes, all my IP's and domains have been edited to protect the innocent.

The VLAN I was having wifi issues with was "LAN", aKa vlan 5

 

[neb50]

Looking for some help. I've been using the brocade6450 as a basic network switch for the past year. Didn't touch any settings at all. Recently upgraded to 3gb internet so wanted to use the 10G SFP slots. I bought two of these from Amazon - https://a.co/d/aFhYnIa

Have this connected to my unraid server which shows a 10000 Mbps eth0 link (previous was 1000 Mbps) so I thought everything was groovy. When I did a speed test though my download and upload fluxuate greatly.


View attachment 33101

This is using a docker tool called speed test tracker does speed tests every 30 mins

My download goes up and down hitting close to max of 2500 (since the router is a 2.5 GB router) but my upload seems to cap below 300.

Wondering if anyone had experience with these trancievers or any thought on what's going on.

The SFP to 10G may not support flow control or something and will cause the packets to drop and get resent.

Try https://www.amazon.com/gp/product/B09ZYGYZ7X/ref=ppx_yo_dt_b_asin_title_o03_s00?ie=UTF8&th=1 or https://www.amazon.com/gp/product/B078SNK1MY/ref=ppx_yo_dt_b_asin_title_o03_s00?ie=UTF8&psc=1

I believe both of those should work but the first one seems to run much cooler.

See this thread for more info. https://forums.servethehome.com/index.php?threads/mix-10g-2-5g-slow-speed-high-retr.33217/

 

[Andydude]

Good luck, I'm putting 3x in a 7250-48 and 1x in a 6450-48.

Anyone know if just 1x MR4020X12B1-RSR in a 6450-48 will cool it ok? It's not PoE and I'm only probably plugging in 7-8 devices...

The 3x MR4020X12B1-RSR in the 725-48 have made it whisper quiet and it runs great, thanks to @RoachedCoach!

However, I put 1x in the 6450-48 and after powering up the fan never slows down from 100% even with nothing plugged in. I only have one power supply and show chassis all looks fine. Am I missing something or does this fan just not work with 6450-48? Anyone got a recommendation for others that are good that I can buy from the UK?

The stack unit 1 chassis info:

Power supply 1 (NA - AC - Regular) present, status ok
Power supply 2 not present

Fan ok, speed (auto): [[1]]<->2

Fan speed switching temperature thresholds:
        1 -> 2 @ 69 deg-C
        1 <- 2 @ 64 deg-C

Sensor B Temperature Readings:
        Current temperature : 28.5 deg-C
Sensor A Temperature Readings:
        Current temperature : 36.0 deg-C
        Warning level.......: 66.0 deg-C
        Shutdown level......: 76.0 deg-C

Thanks!

 

[joku]

Does anyone use cisco AP's on these switches? I just migrated over to this switch, and have my vlans setup, all my other wired devices are working.

The 4 APs are using mobility express, and were working prior to migration. Previous configuration was PVID 4, Tagged 100, 200, 300
Current Brocade setup is dual-mode 1/1/3 to 1/1/6, vlan 100, 200, 300 tagged.

Am I missing something else that allows them to self-discover? Thanks

Very generous offer! Thanks a lot. Today I received another switch I ordered via eBay due to the issues a faced. The serial connection was successful on the first try. So, here is definitely something wrong with the switch itself. I cannot tell you what it is, but it is the switch.

try these baud rates instead, one at a time: 14400, 19200, 38400, 57600, 115200

on the uboot/linux based models like the 6450, it's possible to change the serial baud rate in the u-boot environment variables, I suppose it's possible someone did this previously. I have only ran into it once though, previous user forced it to 115200

Thank you for the tips! I tried all those baud rates one by one, unfortunately without success. I'll return the switch now.

EDIT: @fohdeesha ... And of course, a thousand thanks for this absolutely awesome documentation – really cool stuff! I ran through all the setup/config/licensing steps without any problems.

 

[Testsubg]

Shot in the dark, but I think egiga0 is the Management ethernet port (might be wrong). The error says it's not plugged in. Is that the port you've plugged into the tftp server? Cable good?

I had no issues connecting to my TFTP server using the usb —>serial cable. Also was able to update/flash the new OS update_primary but afterwards trying to update the
bootloader: it connects to the TFTP server but seems to be trying to download and gives # or % (I can’t remember exactly which right now) and then the message about exceeeded number of retry’s allowed. I had powered down both units for a few days and when setting up everything it had the correct image names already, and this could be a remembered setting or listing the versions that are updated. Idk.So I need to update or check version number for uboot and then update the firmware for POE.

the It’s the top left corner. Ive assumed it was the management but if it’s as simple as moving it to the top right I’ll try that when I can this afternoon and report back. Not sure that it would have allowed the primary update if it wasn’t management. Thanks

 

[dreamsin]

I have a spare 6450-24 and 6450-48, but need a 6450-48p... open to trades in EU.

 

[Testsubg]

Just to update this earlier post, the management port is in the top left. And while I repeatedly followed various instructions and was never exactly sure it worked sometimes but after the most recent power cycle and disconnecting the LAN it just looks/presents what I assume is the correct info and versions. Thanks!

 

[fohdeesha]

Just to update this earlier post, the management port is in the top left. And while I repeatedly followed various instructions and was never exactly sure it worked sometimes but after the most recent power cycle and disconnecting the LAN it just looks/presents what I assume is the correct info and versions. Thanks!View attachment 33185

that is an out of date image, and the limited switch-features only image as well. You need to cable the switch properly, then follow the guide from the beginning - the ethernet management port, that you need to plug into your network to update properly, is the BOTTOM left. The top left is the serial console port

 

[richtj99]

Hi,

I updated firmware on my router, noticed some connection issues so I logged into my three ICX-7250's, did a wr mem, then reload. Two of the three of them came up fine but the third one somehow did a factory reset. I have the settings written down and was able to reset it up with no real headaches except - this happened to me before on one of the other switches. I am wondering if I did something wrong or why it factory reset itself?

I am running an older firmware from Fohdeesha. I know I can use a newer firmware but am not really sure if it is something I can upgrade or if it requires factory resetting the switches, updating, then reconfiguring so I have held off on it as things work fine except for the occasional unexpected factory reset.

Any help would be great!
Thanks,
Rich

SSH@Basement#sh ver
  Copyright (c) 2017 Ruckus Wireless, Inc. All rights reserved.
    UNIT 1: compiled on Apr  9 2019 at 03:20:17 labeled as SPR08080e
      (29826604 bytes) from Primary SPR08080e.bin
        SW: Version 08.0.80eT213
      Compressed Boot-Monitor Image size = 786944, Version:10.1.14T215 (spz10114)
       Compiled on Thu Nov 15 08:59:16 2018

  HW: Stackable ICX7250-48-HPOE
==========================================================================
UNIT 1: SL 1: ICX7250-48P POE 48-port Management Module
      Serial  #:DUK2307M0E4
      Software Package: ICX7250_L3_SOFT_PACKAGE   (LID: fwmINFJoFgJ)
      Current License: l3-prem-8X10G
      P-ASIC  0: type B344, rev 01  Chip BCM56344_A0
==========================================================================
UNIT 1: SL 2: ICX7250-SFP-Plus 8-port 80G Module
==========================================================================
 1000 MHz ARM processor ARMv7 88 MHz bus
 8192 KB boot flash memory
 2048 MB code flash memory
 2048 MB DRAM
STACKID 1  system uptime is 11 hour(s) 59 minute(s) 15 second(s)
The system started at 23:08:29 GMT-05 Tue Dec 12 2023

The system : started=warm start   reloaded=by "reload"

 

[ManoftheSea]

I have the settings written down... or if it requires factory resetting the switches, updating, then reconfiguring...

You need to set up a tftp server, then do a "copy running-mem tftp add.res.sof.tftp filename.txt". There's no reason to need things "written down" when you can copy the configuration in digital form, and reload it in digital form, in moments.

I can't explain your "factory reset", but the switches do have two firmware slots, maybe you ended up on the other one somehow?