| Date: 21-FEB-1997 16:49:16.00
From: CRONIC::ANTROBUS "Digital Semiconductor Integrated System Services
21-Feb-1997 1646 -0500"
Subj: Monitoring of hardware RAID arrays with OpenVMS/Digital UNIX/NT
To: HYDRA::NEWMAN
Hi,
In looking through the ASK_SSAG notes file regarding monitoring solutions
for hardware based RAID arrays on NT, VMS and Dunix, I came across your note
entered last October, 1996. We too have embarked on a hardware based
RAID implementation at our site. Our biggest concern now is how do
we know when the array changes state. Were you folks able to cobble together
a solution for this?
regards,
Deb Antrobus
<<< SSAG::DISK$ARCH2:[NOTES$LIBRARY.SSAG]ASK_SSAG.NOTE;32767 >>>
-< Ask the Storage Architecture Group >-
================================================================================
Note 6037.0 notification of change in status of a RAID configuration 1 reply
HYDRA::NEWMAN "Chuck Newman, 508/467-5499 (DTN 297)" 30 lines 3-OCT-1996 08:25
--------------------------------------------------------------------------------
A software vendor wants to know (for both Digital UNIX and OpenVMS) how to
get notified of changes in the status of a RAID configuration. Here is her
reply to my request for more information:
-- Chuck Newman
Thank You for your reply. My question is in regards to storage works raid in
general. Our application monitors and notifies end users of problems that have
or are occuring on their system. Several have started using storage works raid
5. The problem they all have is receiving notification when a copy has
occured due to a hardware loss of a device that is part of the set. They have
lost their hot spare in several instances, and were not aware of the situation,
until another failure has occurred, which results in down time and loss of
data. This is mainly due to the fact that as they were unaware of the auto
copy occurrence they were not able to replace the defective device.
I have now heard about the command console, but need to find out specifics how
we can, as a layered product, on OpenVMS and Digital Unix, discover and report
on errors happening in raid sets. We can currently detect shadow copies, but I
don't believe RAID configurations report similiar errors.
Are there any specific error messages reported to OPCOM (for VMS) or to a log
file (for Digital Unix) that we can try to discover to help our customers
utilize and enjoy the benefits of RAID 5?
best regards,
Carol Rogers
================================================================================
Note 6037.1 notification of change in status of a RAID configuration 1 of 1
KITCHE::schott "Eric R. Schott USG Product Manageme" 18 lines 3-NOV-1996 15:51
-< Needs a script... >-
--------------------------------------------------------------------------------
Hi
There is a script called sys_check, that outputs a lot of data
about Digital UNIX systems. It does go out and grab any hsz data
it can find.
It seems to me that one could use the hsz part of this to get
the HSZ info, and look for empty sparesets, or something in the
failed set.
See http://www-unix.zk3.dec.com/tuning/tools/tools.html
look for sys_check
regards
Eric
|