From: Jeff Barratt (jbarratt@COMPSAT.COM)
Date: Mon Feb 03 2003 - 07:56:26 EST
from:
http://www.storage.ibm.com/ibmsan/products/2109/prod_data/g229-6946.html
"Enterprise-level scalability and high availability when deployed with dual
redundant core/edge fabrics capable of attaching thousands of devices."
This same caveat is not listed for the inrange or mcdata directors rebadged
and sold by IBM. however, dual fabric architecture is generally accepted as
best practice
These are not current (may/june 2002) and of course, don't believe
everything you read. maybe all of this is marketing fluff
http://www.byteandswitch.com/document.asp?doc_id=14773
I am searching for the IBM report where this quote was lifted from:
"While Brocade claims that the 12000 provides high availability, IBM
certification testing has shown that this product does NOT provide the
high-availability capabilities currently provided [by] the IBM-remarketed
McData and Inrange directors," the report says.
As an aside, there is this text in the article, but I haven't heard any
dates:
"The IBM report says it is expecting Brocade to deliver the firmware to
provide high-availability features for the switch sometime in 2003."
http://www.byteandswitch.com/document.asp?doc_id=17665
Don't take this as a slam on the 12000. I wish they had a director class
switch at this time(maybe they do now). Interoperability/managability would
be much easier in homogeneous core-edge fabrics
-----Original Message-----
From: IBM AIX Discussion List [mailto:aix-l@Princeton.EDU]On Behalf Of
Bill Verzal
Sent: Friday, January 31, 2003 2:20 PM
To: aix-l@Princeton.EDU
Subject: Re: System arch
http://www.brocade.com/products/pdf/SW12000_HA_TB_01.pdf
Addressing the Causes of Planned Downtime
To address the most common causes of planned downtime (such as scheduled
maintenance, product upgrades, or the deployment or removal of new
components), the SilkWorm 12000 enables non-disruptive hardware and
software upgrades. To prevent downtime, all major hardware components of
the SilkWorm 12000 switch are field-replaceable and hot-swappable.
Field-replaceable components can be installed without disrupting SAN
operations minimizing downtime for hardware upgrades. To prevent downtime
during software upgrades the SilkWorm 12000 provides hot code downloading
which enables firmware upgrades to be downloaded and activated without
disrupting other operations
Perhaps there is a misunderstanding of "firmware" and "software". The
marketing fluff above says you can do hot code downloading.
Whose right ?
And yes - separate power locations would be wonderful as well.
BV
----------------------------------------------------------------------------
----------------------------
Bill Verzal
Technical Consultant
Forbes Technical Consulting
(312) 653-3684
bill_verzal@bcbsil.com
billverzal@imcingular.com (Pager)
888-428-4025 (Pager)
MailStop: 27.202B
"Jeff Barratt" <jbarratt@COMPSAT.COM>
Sent by: "IBM AIX Discussion List" <aix-l@Princeton.EDU>
01/31/2003 12:55 PM
Please respond to "IBM AIX Discussion List"
To: aix-l@Princeton.EDU
cc:
Subject: Re: System arch
agree with everything, except this:
>>Officially, the Silkworm 12000 is not a director, according to
marketing fluff.
this is not marketing fluff. you can not do concurrent firmware updates
without an interruption of the fabric, hence it is just a big switch, and
is
not director class
>>The racks housing the separate fabrics should be on separate power
feeds.
and preferably in separate locations..if possible
-----Original Message-----
From: IBM AIX Discussion List [mailto:aix-l@Princeton.EDU]On Behalf Of
Bill Verzal
Sent: Friday, January 31, 2003 1:51 PM
To: aix-l@Princeton.EDU
Subject: Re: System arch
Please don't take offense at this, but if this is new to you, you should
not do it without a thorough understanding of SAN fabrics, subsystem
device drivers, and switch zoning. This sounds like a pretty substantial
undertaking.
>From a high level, you'll need several core switches and many edge
switches. If you don't know what that means, you should get some
immediate SAN training.
>From there, for the edge connections, you'll need to know traffic loads
so
you can distribute the load across the different edge switches.
You'll need to know the total load for any number of edge switches so you
don't overload the core switches.
You'll need to plan for growth in the edge switches.
You'll need at least 2 separate fabrics to ensure at least 99.999%
availability. The new Brocade Silkworm 12000 advertises (10) 9's in a
dual fabric configuration. That is 99.99999999% uptime, which is .00005
minutes of downtime per year. You also have the option of other models of
directors. Officially, the Silkworm 12000 is not a director, according to
marketing fluff.
The racks housing the separate fabrics should be on separate power feeds.
I'll stop here and let others share their thoughts.
BV
----------------------------------------------------------------------------
----------------------------
Bill Verzal
Technical Consultant
Forbes Technical Consulting
(312) 653-3684
bill_verzal@bcbsil.com
billverzal@imcingular.com (Pager)
888-428-4025 (Pager)
MailStop: 27.202B
"rohan chandrashekar" <rohanchandrashekar@YAHOO.COM>
Sent by: "IBM AIX Discussion List" <aix-l@Princeton.EDU>
01/31/2003 12:27 PM
Please respond to "IBM AIX Discussion List"
To: aix-l@Princeton.EDU
cc:
Subject: System arch
Hi!
There is a plan to open a new datacentre with about 200-300
servers consisting of Sun/IBM and Windows server. The SAN environment
would be using the Brocade switces, EMC symm, storagetek library,
JNI/Emulex hba's. EMC control centre would be used for storage management.
I have been given the responsibility to plan the lay out the SAN arch.
This is quite new to me. I would appreciate if anyone could give some
feedback on this. Any arch diagram or any documents that could be use to
plan the datacentre.
Thanks,
Rohan
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now
**********
The information contained in this communication is confidential, private,
proprietary, or otherwise privileged and is intended only for the use of
the
addressee. Unauthorized use, disclosure, distribution or copying is
strictly prohibited and may be unlawful. If you have received this
communication in error, please notify the sender immediately at
(312)653-6000 in Illinois; (972)766-6900 in Texas; or (800)835-8699 in New
Mexico.
**********
**********
The information contained in this communication is confidential, private,
proprietary, or otherwise privileged and is intended only for the use of the
addressee. Unauthorized use, disclosure, distribution or copying is
strictly prohibited and may be unlawful. If you have received this
communication in error, please notify the sender immediately at
(312)653-6000 in Illinois; (972)766-6900 in Texas; or (800)835-8699 in New
Mexico.
**********
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 22:16:33 EDT