Fibre Channel (SAN)

Reply
Occasional Contributor
Posts: 7
Registered: ‎08-27-2007

extended fabric problem between two DCX backbones

Hi There

Do anyone had a problem with a data replication between two DCX 's on a extended fabric

Let me describe my problem : I cannot do any data replication between my two DCX's running over a DWDM on 4GB

But if I try to do the same between two 48000 over the same DWDM it works fine . only difference is that the 48000 is 4GB and the DCX is 8 GB ports

Any suggestion if I want to use the DCX without any hickups

New Contributor
Posts: 2
Registered: ‎04-06-2010

Re: extended fabric problem between two DCX backbones

Your switch ports are set to auto negotiate right? Your not hard setting links at 8Gb on each side of the 4gb dwdm? We have dark fiber between our two DC's with in metro distance and just use the 30km single mode optics. Saves a ton on DWDM costs.

-Will

Super Contributor
Posts: 635
Registered: ‎04-12-2010

Re: extended fabric problem between two DCX backbones

Did you have compared your portcfg settings between both platforms?

It could also help if you post some informationof the DWDM solution.

Distance and vendor and so on...

Can you please be a bit more specific what your error is between both DCXes?

Did you have a performance issue or latency issue or did no ISL connection will be establish between both directors.

Regards,

Andreas

Frequent Contributor
Posts: 104
Registered: ‎07-27-2009

Re: extended fabric problem between two DCX backbones

  1. Check fillwords. On 8G they should normally be arb(ff)
  2. Verify if your DWDM kit supports that in addition to virtual channels.
  3. If not, use IDLE fillword, turn off credit-recovery and VC link init and turn on ISL R_RDY mode.
  4. Ask your DWDM vendor for either a software upgrade to support these functions or, if thats not possible, shop around for new kit.


The ARB(ff) fillword doesn't exist on the 48K since thats 4G. (Only in loop mode you might run into ARB(ff) but that doesn't apply here). The ARB(ff) fillword was introduced on 8G speeds again since it provides a better bitpattern to prevent some electrical issues. Google around, enough docs availble on this topic. You might want to start of here : http://www.hitachi-storage.com/content/fill-words-what-are-those-what-do-they-do-and-why-are-they-needed

Good luck

Erwin

Join the Community

Get quick and easy access to valuable resource designed to help you manage your Brocade Network.

vADC is now Pulse Secure
Download FREE NVMe eBook