Here ya Go!
LDEV : 1 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER1)
Device | 16777200 | 9879264 ( 59%) | 6897936 ( 41%) |
Permanent | 12582864 ( 75%) | 7844768 ( 47%) | 4738096 ( 28%) |
Transient | 12582864 ( 75%) | 2034496 ( 12%) | 6897936 ( 41%) |
LDEV : 2 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER2)
Device | 25165808 | 14591312 ( 58%) | 10574496 ( 42%) |
Permanent | 18874352 ( 75%) | 11828272 ( 47%) | 7046080 ( 28%) |
Transient | 18874352 ( 75%) | 2763040 ( 11%) | 10574496 ( 42%) |
LDEV : 3 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER3)
Device | 25165808 | 14562816 ( 58%) | 10602992 ( 42%) |
Permanent | 18874352 ( 75%) | 11828896 ( 47%) | 7045456 ( 28%) |
Transient | 18874352 ( 75%) | 2733920 ( 11%) | 10602992 ( 42%) |
LDEV : 4 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER4)
Device | 25165808 | 14496848 ( 58%) | 10668960 ( 42%) |
Permanent | 18874352 ( 75%) | 12113200 ( 48%) | 6761152 ( 27%) |
Transient | 18874352 ( 75%) | 2383648 ( 9%) | 10668960 ( 42%) |
plus from sysgen
LDEV: 1 DEVNAME: OUTDEV: 0 MODE:
ID: HPDARRAY RSIZE: 128 DEVTYPE: DISC
PATH: 0/8/0/0.6.0 MPETYPE: 4 MPESUBTYPE: 2
CLASS: DISC SPOOL
LDEV: 2 DEVNAME: OUTDEV: 0 MODE:
ID: HPDARRAY RSIZE: 128 DEVTYPE: DISC
PATH: 0/8/0/0.6.1 MPETYPE: 4 MPESUBTYPE: 2
CLASS: DISC SPOOL
LDEV: 3 DEVNAME: OUTDEV: 0 MODE:
ID: HPDARRAY RSIZE: 128 DEVTYPE: DISC
PATH: 0/8/0/0.6.2 MPETYPE: 4 MPESUBTYPE: 2
CLASS: DISC SPOOL
LDEV: 4 DEVNAME: OUTDEV: 0 MODE:
ID: HPDARRAY RSIZE: 128 DEVTYPE: DISC
PATH: 0/8/0/0.6.3 MPETYPE: 4 MPESUBTYPE: 2
CLASS: DISC SPOOL
The 40 18Gb disks are clearly visible and are not encased in AutoRaid or Nike,
hence is must be software. Particularly since dstat all looks like
30-ST318452LC MASTER-MD MEMBER1 set name
31-ST318452LC MEMBER-MD MEMBER2 removed
32-ST318452LC MEMBER-MD MEMBER3 to shield
33-ST318452LC MEMBER-MD MEMBER4 the guilty
34-ST318452LC MEMBER-MD MEMBER5
35-ST318452LC MEMBER-MD MEMBER6
> You have not shown that ldevs 1 thru 4 are indeed behaving in raid 5 mode.
>
> (DISCFREE C output, assuming all six 9gb drives are allocated to 4 luns,
> would demonstrate that)
>
> In any case, assuming that they are operating in raid5 mode, that alone
> could suggest a poor choice for ldev 1.
>
> Generally speaking, RAID5 is a poor choice for optimum performance on a
HP3000.
>
> Of course this is a generalized statement, and, as always, your mileage may
> vary.
>
> With the AUTORAID12H, you can ensure RAID1 operation by configuring 50% or
> less of the available total disk space available among up to 8 luns (or
ldevs).
>
> You did not indicate what comprises your RAID 1 disks?
>
> Software mirroring, AUTORAID, or NIKE?
>
>
> At 03:05 PM 2002-10-30, Paul Courry wrote:
> >System:
> >N-class 550 4 way
> >16Gb RAM
> >main app is Ecometry software
> >40 18Gb drives in a Raid 1 configuration for production data
> >AutoRaid with 6 9Gb drives showing as 4 drives in sysgen (I believe this is a
> >raid 5 config discfree shows
> >
> >-----------------------------------------------------------------------------
--
> >LDEV : 1 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER1)
> >
> >LARGEST FREE AREA: 1192336 TOTAL FREE SPACE: 6720256
> >
> > 0 BLOCK(S) OF 1- 9 CONTIG. SECTORS = 0 FREE
> > SECTORS. 0%
> > 2200 BLOCK(S) OF 10- 99 CONTIG. SECTORS = 95712 FREE
> > SECTORS. 1%
> > 2745 BLOCK(S) OF 100- 999 CONTIG. SECTORS = 809424 FREE
> > SECTORS. 12%
> > 470 BLOCK(S) OF 1000- 9999 CONTIG. SECTORS = 1179648 FREE
> > SECTORS. 18%
> > 57 BLOCK(S) OF 10000- 99999 CONTIG. SECTORS = 1201984 FREE
> > SECTORS. 18%
> > 9 BLOCK(S) OF 100000-AND UP CONTIG. SECTORS = 3433488 FREE
> > SECTORS. 51%
> >
> >-----------------------------------------------------------------------------
--
> >LDEV : 2 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER2)
> >
> >LARGEST FREE AREA: 1665760 TOTAL FREE SPACE: 10096240
> >
> > 0 BLOCK(S) OF 1- 9 CONTIG. SECTORS = 0 FREE
> > SECTORS. 0%
> > 3482 BLOCK(S) OF 10- 99 CONTIG. SECTORS = 152736 FREE
> > SECTORS. 2%
> > 4352 BLOCK(S) OF 100- 999 CONTIG. SECTORS = 1343616 FREE
> > SECTORS. 13%
> > 744 BLOCK(S) OF 1000- 9999 CONTIG. SECTORS = 1750752 FREE
> > SECTORS. 17%
> > 79 BLOCK(S) OF 10000- 99999 CONTIG. SECTORS = 1955120 FREE
> > SECTORS. 19%
> > 13 BLOCK(S) OF 100000-AND UP CONTIG. SECTORS = 4894016 FREE
> > SECTORS. 48%
> >
> >-----------------------------------------------------------------------------
--
> >LDEV : 3 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER3)
> >
> >LARGEST FREE AREA: 1747904 TOTAL FREE SPACE: 10098224
> >
> > 0 BLOCK(S) OF 1- 9 CONTIG. SECTORS = 0 FREE
> > SECTORS. 0%
> > 3260 BLOCK(S) OF 10- 99 CONTIG. SECTORS = 142192 FREE
> > SECTORS. 1%
> > 4482 BLOCK(S) OF 100- 999 CONTIG. SECTORS = 1375792 FREE
> > SECTORS. 14%
> > 752 BLOCK(S) OF 1000- 9999 CONTIG. SECTORS = 1671760 FREE
> > SECTORS. 17%
> > 64 BLOCK(S) OF 10000- 99999 CONTIG. SECTORS = 1454528 FREE
> > SECTORS. 14%
> > 14 BLOCK(S) OF 100000-AND UP CONTIG. SECTORS = 5453952 FREE
> > SECTORS. 54%
> >
> >-----------------------------------------------------------------------------
--
> >LDEV : 4 -- (MPEXL_SYSTEM_VOLUME_SET:MEMBER4)
> >
> >LARGEST FREE AREA: 1902144 TOTAL FREE SPACE: 10159328
> >
> > 0 BLOCK(S) OF 1- 9 CONTIG. SECTORS = 0 FREE
> > SECTORS. 0%
> > 3107 BLOCK(S) OF 10- 99 CONTIG. SECTORS = 137232 FREE
> > SECTORS. 1%
> > 4425 BLOCK(S) OF 100- 999 CONTIG. SECTORS = 1521328 FREE
> > SECTORS. 15%
> > 1169 BLOCK(S) OF 1000- 9999 CONTIG. SECTORS = 2595600 FREE
> > SECTORS. 26%
> > 30 BLOCK(S) OF 10000- 99999 CONTIG. SECTORS = 579216 FREE
> > SECTORS. 6%
> > 14 BLOCK(S) OF 100000-AND UP CONTIG. SECTORS = 5325952 FREE
> > SECTORS. 52%
> >
> >-----------------------------------------------------------------------------
--
> >
> >Would someone care to explain to me (let me play dumb) why Autoraid is a poor
> >choice for the system volume set?
> >
> >Thanks!
> >
> >Paul
> >Courry
> >
> >* To join/leave the list, search archives, change list settings, *
> >* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
>
> ------------------------------------------------------------------------------
-------------------
> Gilles Schipper
> GSA Inc.
> HP System Administration Specialists
> 300 John Street, Box 87651 Thornhill, ON Canada L3T 7R4
> Voice: 905.889.3000 Fax: 905.889.3001
> email: [log in to unmask] web: http://www.gsainc.com
> ------------------------------------------------------------------------------
-------------------
>
> * To join/leave the list, search archives, change list settings, *
> * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
|
