[geeks] Seagate vendor-unique data?

[erie patsellis]

On 2/2/2012 1:43 PM, Jochen Kunz wrote:
> On Thu, 2 Feb 2012 11:19:51 -0600
> Bill Bradford<mrbill at mrbill.net>  wrote:
>
>> I remember reading something quite recently about manually setting drive
>> IDs on (DSSI?) hard drives on a VAX by talking to the drive's built-in
>> controller via command line directly...
> Yup. DEC did a lot of very cool things in its days. DSSI is actually a
> "cheap" version of CI (Cluster Interconnect). A DSSI disk is basicly a
> disk with build in HSC (Hirarchical Storage Controler) and thus just an
> other cluster node. CI / DSSI is more a Storage Area Network then just
> a storage bus like e.g. SCSI. - It was this at a time when the term SAN
> wasn't invented yet.
>
> It was quite common for some VAXen that the peripherals had more
> horsepower then the actual main CPU. E.g. the RQDX3 MFM disk controller
> used in low end QBus VAXen and PDP-11s is driven by a T11. (A PDP-11
> micorcontoler.) IIRC the RA90 disks are controled by a NS32K CPU...
All this talk is bringing back memories of working at DEC-ENO building HSC boards. The 
LO108 (KSDI) boards had 4 bitsliced AMD processors on them, you had to learn the microcode 
before you could do most component level repairs as the microcode usually lead you to the 
symptom, then you had to work backward with a logic analyzer to figure out what *actually* 
failed the test. Given that this was early/mid '80s it was pretty interesting that between 
the custom hybrid ECL/TTL line drivers, the oddball PALs and being limited in board plane 
depth that they even worked at all.