Backup Central

I've recently read the following document, which describes how drives
should be configured in Solaris:

http://sony.storagesupport.com/naslibrary/downloads/unix216.pdf

Thus, I have the following in my st.conf:

"SONY SDZ-100", "Sony Super-AIT1", "SONY_SAIT";
SONY_SAIT = 1, 0x36, 0, 0x1d639, 4, 0x00, 0x00, 0x00, 0x00,0;

When I boot the system, I get the following messages:

Dec 4 14:38:20 britt scsi: [ID 799468 kern.info] st17 at fp0: name
w211f0090a5001194,3, bus address ef
Dec 4 14:38:20 britt genunix: [ID 936769 kern.info] st17 is
/pci < at > 1d,700000/SUNW,jfca < at > 1/fp < at > 0,0/st < at > w211f0090a5001194,3
Dec 4 14:38:20 britt scsi: [ID 107833 kern.notice]
/pci < at > 1d,700000/SUNW,jfca < at > 1/fp < at > 0,0/st < at > w211f0090a5001194,2 (st18):
Dec 4 14:38:20 britt SONY_SAIT options value invalid bits set: 0x4000
Dec 4 14:38:20 britt scsi: [ID 365881 kern.info]
/pci < at > 1d,700000/SUNW,jfca < at > 1/fp < at > 0,0/st < at > w211f0090a5001194,2 (st18):
Dec 4 14:38:20 britt <Sony Super-AIT1>

According to the above document, those bits should be:

ST_BUFFERED_WRITES 0x4000

So, my configuration should be valid..?

Which leads me to my question:

Does anyone have a verified working st.conf for use with S-AIT1 tape
devices? (SDZ-100) I'm running solaris 9, and its possible that the above
document isn't updated.

I'm asking because I get fatal timeouts to the device during some rare
circumstances. I'm not sure exactly what's causing it though.

//Oscar

Note: To sign off this list, send a "signoff networker" command via email
should be sent to stan < at > temple.edu

On Sat, 4 Dec 2004, Oscar Olsson wrote:

I've recently read the following document, which describes how drives
should be configured in Solaris:

http://sony.storagesupport.com/naslibrary/downloads/unix216.pdf

Thus, I have the following in my st.conf:

"SONY SDZ-100", "Sony Super-AIT1", "SONY_SAIT";
SONY_SAIT = 1, 0x36, 0, 0x1d639, 4, 0x00, 0x00, 0x00, 0x00,0;


I have these entries

"SONY SDZ-130", "SONY 1/2 INCH S-AIT", "SONY_SAIT";

and

SONY_SAIT = 1, 0x36, 0, 0x19639, 4, 0x00, 0x00, 0x00, 0x00, 0;

which I copied and pasted directly from my server's st.conf file.
Our S-AIT drives are connected via a fiber hub so there might be
a difference if your drives are connected via traditional SCSI. If
you continue to have a problem, just get on the phone with your
Sony techncal contact and ask him or her to email you the proper
st.conf file entries. My equipment was installed by Sony who came
on-site so I know that what I have has Sony's blessing.

Note: To sign off this list, send a "signoff networker" command via email
should be sent to stan < at > temple.edu

On Sat, 4 Dec 2004, Stan Horwitz wrote:

SH> On Sat, 4 Dec 2004, Oscar Olsson wrote:
SH>
SH> >I've recently read the following document, which describes how drives
SH> >should be configured in Solaris:
SH> >
SH> >http://sony.storagesupport.com/naslibrary/downloads/unix216.pdf
SH> >
SH> >Thus, I have the following in my st.conf:
SH> >
SH> >"SONY SDZ-100", "Sony Super-AIT1", "SONY_SAIT";
SH> >SONY_SAIT = 1, 0x36, 0, 0x1d639, 4, 0x00, 0x00, 0x00, 0x00,0;
SH> >
SH>
SH> I have these entries
SH>
SH> "SONY SDZ-130", "SONY 1/2 INCH S-AIT", "SONY_SAIT";
SH>
SH> and
SH>
SH> SONY_SAIT = 1, 0x36, 0, 0x19639, 4, 0x00, 0x00, 0x00, 0x00, 0;

OK, the difference is 4000 between 1d639 and 19639, so its possible that
it might be an outdated option. I'll check with SpectraLogic, who are
responsible for the Sony drives in the T950 library. I bet I won't get any
support directly from Sony since these are OEM drives behind a spectra
QIP.

Oh, and thanks! :)

//Oscar

Note: To sign off this list, send a "signoff networker" command via email
should be sent to stan < at > temple.edu

On Sat, 4 Dec 2004, Oscar Olsson wrote:

OK, the difference is 4000 between 1d639 and 19639, so its possible that
it might be an outdated option. I'll check with SpectraLogic, who are
responsible for the Sony drives in the T950 library. I bet I won't get any
support directly from Sony since these are OEM drives behind a spectra
QIP.

Oh, you're probably right about contacting Sony for support. My S-AIT
drives are sitting in a Sony PetaSite so I guess the tech support options
you have are different from what I have. When we were shopping for a tape
library last winter, we seriously looked at the T950, but we went with the
PetaSite. We're happy campers with our PetaSite, but the T950 looked very
interesting too. This was a tough decision for us to make. Good luck with
your T950.

Note: To sign off this list, send a "signoff networker" command via email
should be sent to stan < at > temple.edu

Thus, I have the following in my st.conf:

"SONY SDZ-100", "Sony Super-AIT1", "SONY_SAIT";
SONY_SAIT = 1, 0x36, 0, 0x1d639, 4, 0x00, 0x00, 0x00, 0x00,0;

When I boot the system, I get the following messages:

Dec 4 14:38:20 britt scsi: [ID 799468 kern.info] st17 at fp0: name
w211f0090a5001194,3, bus address ef
Dec 4 14:38:20 britt genunix: [ID 936769 kern.info] st17 is
/pci < at > 1d,700000/SUNW,jfca < at > 1/fp < at > 0,0/st < at > w211f0090a5001194,3
Dec 4 14:38:20 britt scsi: [ID 107833 kern.notice]
/pci < at > 1d,700000/SUNW,jfca < at > 1/fp < at > 0,0/st < at > w211f0090a5001194,2 (st18):
Dec 4 14:38:20 britt SONY_SAIT options value invalid bits set: 0x4000
Dec 4 14:38:20 britt scsi: [ID 365881 kern.info]
/pci < at > 1d,700000/SUNW,jfca < at > 1/fp < at > 0,0/st < at > w211f0090a5001194,2 (st18):
Dec 4 14:38:20 britt <Sony Super-AIT1>

According to the above document, those bits should be:

ST_BUFFERED_WRITES 0x4000

So, my configuration should be valid..?

Certain st patches for Solaris 8 and all of Solaris 9 appear to remove
the 0x4000 bit as valid.

Basically, the st driver no longer allows write acknowledgements to be
sent to the application before the write reaches the tape drive. This
could be a slight performance hit, but it removes potential problems if
the buffer is not flushed later.

You shouldn't really notice any problems, just drop the bit from the
line to avoid the warning message.

Does anyone have a verified working st.conf for use with S-AIT1 tape
devices? (SDZ-100) I'm running solaris 9, and its possible that the above
document isn't updated.

I'm asking because I get fatal timeouts to the device during some rare
circumstances. I'm not sure exactly what's causing it though.

The error you show at st load time shouldn't have anything to do with
those timeouts.


Darren Dunham ddunham < at > taos.com
Senior Technical Consultant TAOS http://www.taos.com/
Got some Dr Pepper? San Francisco, CA bay area
< This line left intentionally blank to confuse you. >

Note: To sign off this list, send a "signoff networker" command via email
should be sent to stan < at > temple.edu

On Sun, 5 Dec 2004, Darren Dunham wrote:

DD> Certain st patches for Solaris 8 and all of Solaris 9 appear to remove
DD> the 0x4000 bit as valid.

Yep, figured out that a while after I sent that email. It doesn't complain
at boot anymore.

DD> The error you show at st load time shouldn't have anything to do with
DD> those timeouts.

No, it turned out to be a typo in st.conf - I forgot one set of 0x00 for
the SAIT definition. Since I corrected that, I haven't seen any
timeouts. Solaris probably assumed some incorrect values which the drives
didn't like from time to time, at least that's my guess.

//Oscar

Note: To sign off this list, send a "signoff networker" command via email
should be sent to stan < at > temple.edu