U320 Cable Lengths

Kevin D Butt kdbutt at us.ibm.com
Mon Aug 25 10:28:52 PDT 2003


* From the T10 Reflector (t10 at t10.org), posted by:
* Kevin D Butt <kdbutt at us.ibm.com>
*
This is a multipart message in MIME format.
--=_alternative 005F735707256D8D_=
Content-Type: text/plain; charset="us-ascii"


Paul and Leon, 

Thanks for your responses, this gives me a lot to think about.  It
appears that your notes relate specifically to HDD and RAID type of
devices.  Do you have any experience related to external cabling? 

Thanks,

Kevin D. Butt
Fibre Channel & SCSI Architect, IBM Tape Microcode, 
6TYA, 9000 S. Rita Rd., Tucson, AZ  85744
Tie-line 321; Office: 520-799-5280, Lab: 799-5751, Fax: 799-4138, Email:
kdbutt at us.ibm.com 



	"Aloisi, Paul" <paul_aloisi at ti.com> 
Sent by: owner-t10 at t10.org 


08/25/2003 03:56 AM 
        
        To:        "'Leon Toh'" <tltoh at attglobal.net>, t10 at t10.org 
        cc:         
        Subject:        RE: U320 Cable Lengths 





* From the T10 Reflector (t10 at t10.org), posted by:
* "Aloisi, Paul" <paul_aloisi at ti.com>
*
Ultra320 SCSI requires different cables and spacing than Ultra160 or
Ultra2
SCSI. Unfortunately the PIP and SSM standards were not ready when
Ultra320
was released. The successful multidrop cables that I have used for
Ultra320
and Ultra640 testing were based on 5 1/4 inch spacing between flats with
the
connectors placed at 2 or 3 flats apart, alternating the spacing. The
9.85
inch spacing used for Ultra160 and Ultra2 are actually a filter for the
Ultra320 and Ultra640 speeds. Most backplanes had to be redesigned for
different spacing, many of the backplanes had the same problem as the
cables, creating a filter for the Ultra320 and Ultra640 speeds.

What helps a lot of these system is to drop the terminator impedance, if
you
used the SPI-5 terminator for Ultra320 systems and drop the impedance to
minimize the reflection from the terminator, this will help a lot on
backplanes with close spacing. 

The SPI-5 standard refers to both PIP and SSM-2 standards, these are all
published standards. If you used the SPI-5 documentation for Ultra160,
Ultra320 and Ultra640 systems, you will get much better results.
Unfortunately the programmable terminator, PIP and SSM-2 were not ready
when
we released SPI-4, which first documents the Ultra320 speed. 

Thank you,
Paul Aloisi

-----Original Message-----
From: Leon Toh [mailto:tltoh at attglobal.net] 
Sent: Monday, August 25, 2003 2:03 AM
To: t10 at t10.org
Subject: Re: U320 Cable Lengths

* From the T10 Reflector (t10 at t10.org), posted by:
* Leon Toh <tltoh at attglobal.net>
*
U320 have been a real nightmare and it still is for many of us. My 
experience here are mostly with U320 HDD in RAID environment.

Once I have 4 or more U320 HDD attach to a multi-drop cable I start
having 
nightmares. At most times with 4 U320 HDD I find that the HBA will 
negotiate reliably for most of the HDD at U160. If I wanted U320 
reliability than I have to take a number of factors into consideration.
For 
me I find that the type of multi-drop cable and cable quality happens to

play a major role besides HDD firmware when comes to U320 negotiation 
reliability.

As you mention that the specs specifies that for a multi-drop
configuration 
maximum allowed cable length is 12.5 m. To me this is virtually
impossible 
going by the number of issues which I have encountered. By the way we
are 
also having issues when comes to hotswap passive backplane using U320
HDD's 
also.

Majority of the issues I'm experiencing here I would say they are mostly

noise related. Reason I got to this conclusion was mainly the fact that 
there is no consistency at all and in some instances using the same 
multi-drop cable, HDD and HBA but install into a completely different 
Server / Workstation sometimes it work reliably for me.

I hope that HDD vendor's won't be releasing a U640 parallel version when

the specs is finally finalise. Otherwise I can imagine the amount of
issues 
we would face when parallel U640 version becomes available.


Best Regards,
Leon

At 09:25 AM 23/08/03, Kevin D Butt wrote:

>I have a question about real-world cable lengths in use in Ultra 320 
>environments.  SPI-5 lists the cable lengths at 25 m for point-to-point

>and 12,5 m for multi-drop.
>Is there anybody out there who is actually using 25 m cables in point
to 
>point without any issues?
>Is there anybody out there using 12,5 m cables with mulitple devices 
>attached without any issues?
>How many devices have successfully been used on one cable?
>
>Thanks,
>
>Kevin D. Butt
>Fibre Channel & SCSI Architect, IBM Tape Microcode,
>6TYA, 9000 S. Rita Rd., Tucson, AZ  85744
>Tie-line 321; Office: 520-799-5280, Lab: 799-5751, Fax: 799-4138,
Email: 
>kdbutt at us.ibm.com

*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org




--=_alternative 005F735707256D8D_=
Content-Type: text/html; charset="us-ascii"


<br><font size=2 face="sans-serif">Paul and Leon,</font>
<br>
<br><font size=2 face="sans-serif">Thanks for your responses, this gives me a lot to think about. &nbsp;It appears that your notes relate specifically to HDD and RAID type of devices. &nbsp;Do you have any experience related to external cabling?</font>
<br>
<br><font size=2 face="sans-serif">Thanks,<br>
<br>
Kevin D. Butt<br>
Fibre Channel &amp; SCSI Architect, IBM Tape Microcode, <br>
6TYA, 9000 S. Rita Rd., Tucson, AZ &nbsp;85744<br>
Tie-line 321; Office: 520-799-5280, Lab: 799-5751, Fax: 799-4138, Email: kdbutt at us.ibm.com</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td>
<td><font size=1 face="sans-serif"><b>"Aloisi, Paul" <paul_aloisi at ti.com&gt;</b></font>
<br><font size=1 face="sans-serif">Sent by: owner-t10 at t10.org</font>
<p><font size=1 face="sans-serif">08/25/2003 03:56 AM</font>
<br>
<td><font size=1 face="Arial">&nbsp; &nbsp; &nbsp; &nbsp; </font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; To: &nbsp; &nbsp; &nbsp; &nbsp;"'Leon Toh'" <tltoh at attglobal.net&gt;, t10 at t10.org</font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; cc: &nbsp; &nbsp; &nbsp; &nbsp;</font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; Subject: &nbsp; &nbsp; &nbsp; &nbsp;RE: U320 Cable Lengths</font>
<br></table>
<br>
<br>
<br><font size=2 face="Courier New">* From the T10 Reflector (t10 at t10.org), posted by:<br>
* "Aloisi, Paul" <paul_aloisi at ti.com&gt;<br>
*<br>
Ultra320 SCSI requires different cables and spacing than Ultra160 or Ultra2<br>
SCSI. Unfortunately the PIP and SSM standards were not ready when Ultra320<br>
was released. The successful multidrop cables that I have used for Ultra320<br>
and Ultra640 testing were based on 5 1/4 inch spacing between flats with the<br>
connectors placed at 2 or 3 flats apart, alternating the spacing. The 9.85<br>
inch spacing used for Ultra160 and Ultra2 are actually a filter for the<br>
Ultra320 and Ultra640 speeds. Most backplanes had to be redesigned for<br>
different spacing, many of the backplanes had the same problem as the<br>
cables, creating a filter for the Ultra320 and Ultra640 speeds.<br>
<br>
What helps a lot of these system is to drop the terminator impedance, if you<br>
used the SPI-5 terminator for Ultra320 systems and drop the impedance to<br>
minimize the reflection from the terminator, this will help a lot on<br>
backplanes with close spacing. <br>
<br>
The SPI-5 standard refers to both PIP and SSM-2 standards, these are all<br>
published standards. If you used the SPI-5 documentation for Ultra160,<br>
Ultra320 and Ultra640 systems, you will get much better results.<br>
Unfortunately the programmable terminator, PIP and SSM-2 were not ready when<br>
we released SPI-4, which first documents the Ultra320 speed. <br>
<br>
Thank you,<br>
Paul Aloisi<br>
<br>
-----Original Message-----<br>
From: Leon Toh [mailto:tltoh at attglobal.net] <br>
Sent: Monday, August 25, 2003 2:03 AM<br>
To: t10 at t10.org<br>
Subject: Re: U320 Cable Lengths<br>
<br>
* From the T10 Reflector (t10 at t10.org), posted by:<br>
* Leon Toh <tltoh at attglobal.net&gt;<br>
*<br>
U320 have been a real nightmare and it still is for many of us. My <br>
experience here are mostly with U320 HDD in RAID environment.<br>
<br>
Once I have 4 or more U320 HDD attach to a multi-drop cable I start having <br>
nightmares. At most times with 4 U320 HDD I find that the HBA will <br>
negotiate reliably for most of the HDD at U160. If I wanted U320 <br>
reliability than I have to take a number of factors into consideration. For <br>
me I find that the type of multi-drop cable and cable quality happens to <br>
play a major role besides HDD firmware when comes to U320 negotiation <br>
reliability.<br>
<br>
As you mention that the specs specifies that for a multi-drop configuration <br>
maximum allowed cable length is 12.5 m. To me this is virtually impossible <br>
going by the number of issues which I have encountered. By the way we are <br>
also having issues when comes to hotswap passive backplane using U320 HDD's <br>
also.<br>
<br>
Majority of the issues I'm experiencing here I would say they are mostly <br>
noise related. Reason I got to this conclusion was mainly the fact that <br>
there is no consistency at all and in some instances using the same <br>
multi-drop cable, HDD and HBA but install into a completely different <br>
Server / Workstation sometimes it work reliably for me.<br>
<br>
I hope that HDD vendor's won't be releasing a U640 parallel version when <br>
the specs is finally finalise. Otherwise I can imagine the amount of issues <br>
we would face when parallel U640 version becomes available.<br>
<br>
<br>
Best Regards,<br>
Leon<br>
<br>
At 09:25 AM 23/08/03, Kevin D Butt wrote:<br>
<br>
>I have a question about real-world cable lengths in use in Ultra 320 <br>
>environments. &nbsp;SPI-5 lists the cable lengths at 25 m for point-to-point <br>
>and 12,5 m for multi-drop.<br>
>Is there anybody out there who is actually using 25 m cables in point to <br>
>point without any issues?<br>
>Is there anybody out there using 12,5 m cables with mulitple devices <br>
>attached without any issues?<br>
>How many devices have successfully been used on one cable?<br>
><br>
>Thanks,<br>
><br>
>Kevin D. Butt<br>
>Fibre Channel &amp; SCSI Architect, IBM Tape Microcode,<br>
>6TYA, 9000 S. Rita Rd., Tucson, AZ &nbsp;85744<br>
>Tie-line 321; Office: 520-799-5280, Lab: 799-5751, Fax: 799-4138, Email: <br>
>kdbutt at us.ibm.com<br>
<br>
*<br>
* For T10 Reflector information, send a message with<br>
* 'info t10' (no quotes) in the message body to majordomo at t10.org<br>
*<br>
* For T10 Reflector information, send a message with<br>
* 'info t10' (no quotes) in the message body to majordomo at t10.org<br>
</font>
<br>
<br>
--=_alternative 005F735707256D8D_=--




More information about the T10 mailing list