T.R | Title | User | Personal Name | Date | Lines |
---|
6715.1 | | SUBSYS::VIDIOT::PATENAUDE | Ask your boss for ARRAY's... | Thu May 22 1997 13:24 | 14 |
| RZ29B drives do NOT spin down by themselves! Period.
Unless that are braindead or crashed. Period.
ALL RZ products will spin down if TOLD to, but we have no PC GREEN power saver
features that make them spin down on their own.
This is an "urban legend" started because the RZ29B "had" (in it's original
firmware now 3 times removed) a mode page setting refered to as "Power
Management". This feature when enabled would ONLY turn off read/write logic
after 500ms of inactivity and IMMEDIATLY (IE: Milliseconds) turn it back on if
it had work to do. It never spun the drive down. It never took the drive off the
bus.
roger.
|
6715.2 | Yup we don't do that the drive must be broken look at errror logs | SSDEVO::RMCLEAN | | Thu May 22 1997 14:30 | 3 |
| Right we do NOT spindown while it is ONLINE. MSCP and Failedset drives
may be spundown but they should show up in an obvious state and they will
NOT be ONLINE to the host.
|
6715.3 | | netbk2.shr.dec.com::Peters | | Thu May 22 1997 15:32 | 6 |
|
SWCC will detect a drive spun down or any other state change ( broken ... ).
Steve P.
|
6715.4 | only know what I was told.... | FIREBL::LEEDS | From VAXinated to Alphaholic | Wed May 28 1997 14:30 | 20 |
|
>RZ29B drives do NOT spin down by themselves! Period.
>Unless that are braindead or crashed. Period.
Well - it may have been braindead or crashed, but it WAS spun down, and it
took the customer with Mission Critical MCS support 6 hours to figure this
out !!! The "fix" was to replace the drive, but their complaint was not
from the fact that the drive "died", but that there was NO indication in the
error log, opcom, HSJ messages, etc. that there was any problem with the
drive. OpenVMS just could not get to certain data, and they finally tracked
the problem down to this drive by "touching" each drive to feel vibration.
I was not there, I didn't see any of this, but the customer was insistent
that this was the way things happened. Thus the question about how to detect
"spun down" disks ... looks like SWCC *may* do this, if in fact there was
any kind of error issued by the disk when it went away - but since the
OpenVMS errorlog had no entry, I'm not sure SWCC would either...
Arlan
|
6715.5 | | AMCFAC::RABAHY | dtn 471-5160, outside 1-810-347-5160 | Wed May 28 1997 15:18 | 5 |
| If it were a member of a host-based volume shadow set then it would have had
error reports a plenty.
POLYCENTER System Watchdog can poll for just about anything and generate an
error report when something is amiss.
|
6715.6 | | SUBSYS::VIDIOT::PATENAUDE | Ask your boss for ARRAY's... | Wed May 28 1997 21:32 | 9 |
|
re-.2
John, don't take offense. My point was that the drive had to be broken
and not just spun down. The big problem is no errors. At a minimum you
should have gotten a E0 and B0 error for the command timeout and lost
PTL and a FAULT lamp if the drive fell from the face of the earth.
Roger.
|