From: arne@vajhoej.dk
On 12/1/2025 6:56 PM, Rich Alderson wrote:
> =?UTF-8?Q?Arne_Vajh=C3=B8j?= writes:
>> On 12/1/2025 1:02 PM, Scott Dorsey wrote:
>>> There are many devices that should be preserved in working condition so
that
>>> new generations should see how they worked when they were new and
experience
>>> them working. The RA81 should be left in the rack broken, so that new
>>> generations can see how they were when they were new.
>
>> RA81's and RA82's on a 8650.
>
>> The RA82's were okay, but the RA81's crashed frequently. DEC field
>> service knew that site very well.
>
>> I have been told here that there were a manufacturing problem with
>> RA81's and when the problem got fixed they stopped crashing all the
>> time.
>
> You've got the RA81 and RA82 backwards. We had RA81s on a CI cluster at
> Stanford, and they were pretty solid.
>
> RA82s were not available for the 36 bit systems, but we heard about them from
> our DEC FEs.
I could easily remember it wrong - it has been a long time.
But a search in c.o.v confirmed that it was the RA81's.
2003 post:
The problem with the RA81's wasn't that the drives were really flaky --
there were some early data and noise problems ECO's out under warranty
early on -- but that the glue that held the HDA filter was substituted
without a close look at the specifications of the glue.
The substituted glue on HDA's at Rev F (IIRC) would liquify at warm
temperatures and flow to the bottom of the HDA and crash head 13
into the bottom platter every time.
2021 post:
The early RA81s had a problem where the heads would "unglue" from the
actuator and totally trash the disk.
DEC actively hunted out and replaced those original drives. The
replacement RA81s were good and ran forever.
Arne
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
