EGADS!!! Frank Girard's configuration looks very much like one of mine!
We have a system on 6.0 Express 1 plus some patches with ONLY a single
AUTORAID.
The AUTORAID is set up as:
LDEV Size Use
1 4 Gb MPE_voleset
2 6 MPE_colset
72-75
Performance is as good as the same system when on a collection of
internal and Jamaica discs on three FW-SCSI channels.
We are waiting for a software patch and/or a firmware fix that should
improve performance about 10%.
The AUTORAID cache sometimes gives very high I/O rates that can result
in much improved performance.
The AUTORAID handles the Raid 1 and 5 balance automagically, so there is
no direct config of those attributes.
About all one can do is ensure sufficient disc is available (outside of
configured LUNs) so the AUTORAID can keep as much as needed in Raid 1.
So far the balance on out system is very good. But it takes time for
the data to migrate to the mode where it should eventually stay. Note
we do not have large databases but do have 468,000 smallish files.
James (Feeling much safer about disc failures) Overman HP CS R&D
-----Original Message-----
From: Girard, Frank <[log in to unmask]>
Newsgroups: comp.sys.hp.mpe
To: [log in to unmask] <[log in to unmask]>
Date: Wednesday, March 22, 2000 12:40 PM
Subject: Re: 12h disc array
>I can confirm this. When we upgraded our box to a 929x about 6 months
ago,
>We partitioned the drive like that, then installed the OS on ldev1,
which
>was one of the luns of the array. As of last month we are using the
>internal original ldev1 for other uses. Here is a discfree c output:
>
>MP4TEST:DISCFREE C
>
>Ldevs 1-5 are the array, the rest are the old drives from the other
system.
>only 1 drive is not shown, the original ldev1, It is in a loaner state
>waiting to be configured and used for online backup.
>
>DISCFREE A.50.01 Copyright (C) Hewlett-Packard 1992. All rights
reserved.
> WED, MAR 22, 2000, 3:37 PM
>
>ALL MEASUREMENTS ARE IN SECTORS.
>ALL PERCENTAGES ARE RELATIVE TO THE DEVICE SIZE.
>
> | Configured | In Use | Available
|
>-----------+-------------------+-------------------+-------------------
+
>
>LDEV : 1 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER1)
> Device | 16777200 | 4969504 ( 30%) | 11807696 ( 70%)
|
> Permanent | 12582864 ( 75%) | 3629872 ( 22%) | 8952992 ( 53%)
|
> Transient | 12582864 ( 75%) | 1339632 ( 8%) | 11243232 ( 67%)
|
>
>LDEV : 2 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER2)
> Device | 16777200 | 5009872 ( 30%) | 11767328 ( 70%)
|
> Permanent | 12582896 ( 75%) | 3691920 ( 22%) | 8890976 ( 53%)
|
> Transient | 12582896 ( 75%) | 1317952 ( 8%) | 11264944 ( 67%)
|
>
>LDEV : 3 -- (SGA_APPL_VOLUME_SET:MEMBER1)
> Device | 50331632 | 36346096 ( 72%) | 13985536 ( 28%)
|
> Permanent | 50331632 (100%) | 36346096 ( 72%) | 13985536 ( 28%)
|
> Transient | 50331632 (100%) | 0 ( 0%) | 13985536 ( 28%)
|
>
>LDEV : 4 -- (SGA_APPL_VOLUME_SET:MEMBER2)
> Device | 50331632 | 36741904 ( 73%) | 13589728 ( 27%)
|
> Permanent | 50331632 (100%) | 36741904 ( 73%) | 13589728 ( 27%)
|
> Transient | 50331632 (100%) | 0 ( 0%) | 13589728 ( 27%)
|
>
>LDEV : 5 -- (SGA_APPL_VOLUME_SET:MEMBER3)
> Device | 50331632 | 37150032 ( 74%) | 13181600 ( 26%)
|
> Permanent | 50331632 (100%) | 37150032 ( 74%) | 13181600 ( 26%)
|
> Transient | 50331632 (100%) | 0 ( 0%) | 13181600 ( 26%)
|
>
>LDEV : 11 -- (SGA2_APPL_VOLUME_SET:MEMBER1)
> Device | 8330528 | 5010432 ( 60%) | 3320096 ( 40%)
|
> Permanent | 8330528 (100%) | 5010432 ( 60%) | 3320096 ( 40%)
|
> Transient | 8330528 (100%) | 0 ( 0%) | 3320096 ( 40%)
|
>
>LDEV : 12 -- (SGA2_APPL_VOLUME_SET:MEMBER2)
> Device | 8330528 | 5101680 ( 61%) | 3228848 ( 39%)
|
> Permanent | 8330528 (100%) | 5101680 ( 61%) | 3228848 ( 39%)
|
> Transient | 8330528 (100%) | 0 ( 0%) | 3228848 ( 39%)
|
>
>LDEV : 13 -- (SGA2_APPL_VOLUME_SET:MEMBER3)
> Device | 8330528 | 5140160 ( 62%) | 3190368 ( 38%)
|
> Permanent | 8330528 (100%) | 5140160 ( 62%) | 3190368 ( 38%)
|
> Transient | 8330528 (100%) | 0 ( 0%) | 3190368 ( 38%)
|
>
>LDEV : 14 -- (SGA2_APPL_VOLUME_SET:MEMBER4)
> Device | 8330528 | 5148352 ( 62%) | 3182176 ( 38%)
|
> Permanent | 8330528 (100%) | 5148352 ( 62%) | 3182176 ( 38%)
|
> Transient | 8330528 (100%) | 0 ( 0%) | 3182176 ( 38%)
|
>
>TOTALS :
> Device | 217871408 | 140618032 ( 65%) | 77253376 ( 35%)
|
> Permanent | 209482768 ( 96%) | 137960448 ( 63%) | 71522320 ( 33%)
|
> Transient | 209482768 ( 96%) | 2657584 ( 1%) | 76186528 ( 35%)
|
>
>-----Original Message-----
>From: Lars Appel [mailto:[log in to unmask]]
>Sent: Wednesday, March 22, 2000 3:20 PM
>To: [log in to unmask]
>Subject: Re: 12h disc array
>
>
>While I am far from being an AutoRAID expert...
>
>>The 12h (AutoRAID) is supported as of 5.5 pp7, it can be used as the
system
>>volume, and a 9gb drive can be configured as ldev 1 but will only
address
>>up to 4gb.
>
>I was under the impression that the disc sizes inside the AutoRAID are
>kind of "invisible" to the user, er, system side i.e. you configure
some
>LUNs as "slices from the available disc space pool" and these are made
>"visible" on the (FWD) SCSI bus (and become LDEVs then).
>
>This way it should(?) be possible to configure a LUN of max 4 GB for
LDEV
1.
>
>Can anybody with more experience confirm or deny this impression/guess?
>
>
>>The performance implications of the storage system are another story.
As
>>with all things performance, "it depends" is again the right answer.
(...)
>
>And here comes my next "educated guess"... Wouldn't the amount of space
>you allocate for LUNs (compared to the raw disc space that is plugged
in
>the AutoRAID) control how much storage is kept in RAID-1 mode and how
>much has to be migrated to RAID-5 mode?
>
>My impression was something like "plug in 2x N GB disc space, plus some
>add-on for AutoRAIDs internal consumption, hot spare space, etc... and
>then allocate only N GB or less to LUNs, to give the box a chance to
keep
>most of the data in RAID-1 level".
>
>The more space is allocated, the more data will have to be migrated to
>RAID-5 type storage by the array, however it will(?) try to keep most
>frequently accessed data in RAID-1 and less frequently accessed data in
>RAID-5 regions.
>
>Lars "again, not claiming to be an AutoRAID expert"
>
|
