Does SAN 'B' switch Fabric s/w V3.2.x change long distance port matrix

[Malcolm Wade]

I'm planning a migration from 1Gb to 2Gb inter-site links and have been looking at Volume 4 of the SAN Design Reference Guide, page 215 to confirm correct setting for portcfglongdistance on my 'B' style or Brocade switches.

My 2Gb switches are running v3.2.0 firmware which, with the exception of a v3.2.0a release is the latest version and it's been that way for a while yet Table 64 only refers to v3.1 firmware.

Given the differences between v3.0 and v3.1 at 2Gb speeds, can anyone shed any light on wether the matrix has changed further for v3.2 firmware?

Thanks,
Malcolm

[Ruslan]

Hi Malcolm!

How long distance between your switches? There are many factors have an influence on ISL speed, not only firmware release. Explain your SAN design...

[Malcolm Wade]

Ruslan,

I'm upgrading existing 1Gb inter-site links to 2Gb using same fibre runs, just different WDM kit. Fibre run is around the 10 to 15km mark. This setup has been in place for 3+ years with no issues yet really I need to have Extended Fabric enabled if I want to run 2Gb at 5km+ and certainly above 10km and below 25km.

With that in mind I need to set the port that is connecting to my WDM kit to L0.5 which in turn imposes certain 'Quad' rules.

The Design Guide only shows v3.1 rules; I'm running V3.2.

My Fabrics are a mix of 1Gb and 2GB switches (2Gb at the core).

Thanks,
Malcolm

[Peter Mattei · Accepted solution]

Malcolm

The tables are still true for 3.2.x

The only thing that has changed in the long distance arena is the fact that you now can have trunked ports of > 5km!
In the release note it says:

Long-Distance Enhancements

Trunking over extended fabrics (SilkWorm 3200, 3250, 3850, 3800, 3900, 12000, 24000 (all Bloom-ASIC-based platforms) is only supported at 2 Gbit/sec speed, as follows:

- four links at 10 km @ 2 Gbit/sec per trunk group
- three links at 25 km @ 2 Gbit/sec per trunk group
- two links at 50 km @ 2 Gbit/sec per trunk group

When configuring an Extended Fabrics trunk, make sure that the vc_translation_link_init option is set to 1 (enabled). This option can be set using the portCfgLongDistance command. If this option is disabled on a port, the port will not participate in a trunk; instead, it will come up as an individual Lx_Port.

Cheers
Peter

[Ruslan]

Note!

If the fabric contains a mix of switches, use the configure command to make sure the fabric-wide configuration parameter fabric.ops.mode.longDistance is set to 0 on all switches in the fabric.
If the fabric contains SilkWorm 2000-series extended ISLs, use the switchdisable command to disable the switch and then use the configure command to set the fabric-wide configuration parameter fabric.ops.mode.longDistance to 1 on all switches in the fabric. The SilkWorm 4100
model cannot be part of such a fabric.

I highly recommend you read Brocade Fabric OS Procedures Guide. There is many useful tips.

[Malcolm Wade]

Peter and Ruslan,

Thanks for your comments.

Seems I am in the clear; in my current fabric we didn't turn on the Fabric Long Distance Bit for my 1Gb link and by reading the guide and your comments I'll be migrating a 1Gb link to a 2Gb link and setting the new link as L0.5 so I should be ok.

Ruslan mentions the Brocade OS Guide? Anyone have a ilnk to that for download?

Thanks,
Malcolm

[Uwe Zessin]

It's on this page:
http://h20000.www2.hp.com/bizsupport/TechSupport/DocumentIndex.jsp?contentType=SupportManual&locale=en_US&docIndexId=179911&taskId=101&prodTypeId=12169&prodSeriesId=315094

[Ruslan]

Malcolm,

register on brocadeconnect.com site if you really interesting about B-series HP SAN switches

[Malcolm Wade]

My migration to 2Gb inter-site links went smoothly, set the ports to L0.5 and away we went.

[Malcolm Wade]

Closed now it's all in and running.