Disk Enclosures
cancel
Showing results for 
Search instead for 
Did you mean: 

Oracle on an HP AutoRAID Model 12H, HP-UX 11i

SOLVED
Go to solution
Betty Fessler
Frequent Advisor

Oracle on an HP AutoRAID Model 12H, HP-UX 11i

If I have allocated 45GB to one LUN for an oracle
db and I have 50GB left unallocated will it ever revert
to RAID 5 writes
12 REPLIES
Eugeny Brychkov
Honored Contributor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

From your question I see you're expecting answer 'no', but exact answer can be provided only by firmware developer :o)
BTW, do you have latest AutoRAID firmware (HP62)?
Eugeny
Betty Fessler
Frequent Advisor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

HP56 is current f/w
Betty Fessler
Frequent Advisor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

I deleted and re-created the
LUN already, is this enough for the autoRAID to start with
0/1 processing or do I need to
reformat the drives. I'm confused by the Raid 0/1 capacity of 4MB in the arraydsp -a info, when I've
defined 50GB of unallocated
space and the manual states
it will use that for raid 0/1
processing.
Eugeny Brychkov
Honored Contributor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

If it's stated in manual then it should be so. 4MB looks like cache-memory-area-like size than disk area size... post arraydsp -a here
Eugeny
Betty Fessler
Frequent Advisor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

Vendor ID = HP
Product ID = C5447A
Array serial number = 00000037E07C
----------------------------------------------------
Array State = READY
Server name = tepdb
Array type = 3
Mfg. Product Code = IJMTU00003

--- Disk space usage --------------------
Total physical = 138932 MB *
Allocated to LUNs = 46080 MB *
Used as Active Hot spare = 17366 MB *
Used by non-included disks = 0 MB *
Used for Redundancy = 25422 MB *
Unallocated (avail for LUNs) = 50064 MB *
-----------------------------------------


Vendor ID = HP
Product ID = C5447A
Array serial number = 00000037E07C
----------------------------------------------------

LUN 0:
Capacity (MB=2^20) = 46080 MB *
Maximum block addr = 94371840
Block length = 512

Total capacity allocated to LUNs = 46080 MB *----------------------------------------------------
Overall State of Array = READY
Array configuration:
Active Hot Spare Desired = ENABLED
Auto Include = ENABLED
Auto Rebuild = ENABLED
Rebuild Priority = HIGH
Capacity Depletion Threshold = 0%
Write Working Set Interval = 8640 seconds
Language = ENGLISH
Log Full Warning = DISABLED
Volume Set Partitioning = DISABLED
Format Pattern Fill = DISABLED
Subsystem Type ID = 12
LUN Creation Limit = 8
Maximum LUN Creation Limit = 8
Array SCSI configuration:
Controller X SCSI Address = 1
Controller Y SCSI Address = 0
Write Cache = ENABLED
Read Cache = DISABLED
SCSI Parity Checking = ENABLED
SDTR = ENABLED
WDTR = ENABLED
Terminator Power = ENABLED
Unit Attention = ENABLED
Disable Remote Reset = ENABLED
Secondary Controller Offline = DISABLED
Very Early Busy = DISABLED
Queue Full Threshold = 1952
Maximum Queue Full Threshold = 1952
Simplified Resiliency Setting = Unknown
Single Controller Warning = ENABLED
Lock Write Cache On = TRUE
Disable NVRAM on WCE False = FALSE
Disable NVRAM with One Ctrlr = TRUE
Disable NVRAM on UPS absent = FALSE
Force Unit Access Response = 2
Disable Read Hits = FALSE
Resiliency Threshold = 4

Vendor ID = HP
Product ID = C5447A
Array serial number = 00000037E07C
----------------------------------------------------
Raid 0/1 blocks = 8192
Raid 0/1 block length = 512
Raid 0/1 capacity = 4 MB *
Eugeny Brychkov
Honored Contributor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

Rebuild Priority = HIGH
Force Unit Access Response = 2
Raid 0/1 blocks = 8192
Raid 0/1 block length = 512
Raid 0/1 capacity = 4 MB *

Why read cache is disabled? Why rebuild priority is high (will slow down normal host activities in case of rebuild)?
I found doc stating that all large transfers always go to RAID 5 directly and then if needed will migrate to RAID 0/1. In addition, in case of rebuild data will go do RAID 5 also.
So there are only 2 possibilities: either you had rebuild recently or you did large data transfer
Eugeny
Betty Fessler
Frequent Advisor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

This is a used autoraid and the settings are what was there when it came up. And there's not much in the book
about what default settings
are and what it would mean
to change the settings, or how to do it. The book is from Feb 1999. That combined with the
fact that I'm fairly new at this and have never worked with an autoraid means I need
help. I'm trying to set this
up for a whole new oracle db
system, current db size is <10GB but we need some room to
grow. DBA thought maybe 30GB
would cover growth, I decided
to give him 45GB which leaves
the 50GB unallocated. We haven't even loaded the DB yet
which is why I'm confused about the raid 0/1 capacity=
numbers. Having deleted and
recreated the lun it should
show 50gb available for raid 0/1 or maybe that's not what
capacity means in this instance.
Eugeny Brychkov
Honored Contributor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

If you just created a LUN do not expect it will have anything in RAID 0/1. Nothing. You need to write something there, and access it, and you will see that this value will increase
Eugeny
PS. Do you think not assigning points to answers will encourage people to help you in your problems?
Betty Fessler
Frequent Advisor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

Ok, so that raid 0/1 capacity
is whats already used? And
I have been assigning points,
I was just having problems
getting them to stick.
Betty Fessler
Frequent Advisor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

And I just assigned points again and when I went back in they all disappeared again.
Bob_Vance
Esteemed Contributor
Solution

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

The answer to your question is "no". You only have 45GB allocated to LUNs and have 50GB unallocated. Therefore, the 46GB will *always* all be in 1/0 (unless you start creating more LUNs :). Note, that, technically speaking, no space is actually taken up until you actually write some data. So, *theoretically*, you could have more than 50% of the space "allocated" to LUNs and still have all the data in 1/0, simply because not all has been written yet. Of course, this could be a temporary situation.

Also, just to be clear, the 10% limit is not based on the "theory that 90% of the data is unchanged". It's merely to *make sure* that *at least* 10% of the data will be in 1/0. If your active write set is greater than 10% and have no unallocated space, then the array will be forced into updating in 5. The 10% is a minimum cushion so that the situation will never arise that all data is in 5.

Also, the idea of multiple LUNs is further based on the fact the there is an I/O queue in the kernel for each LUN. If you only have one LUN, then all I/O to the array will be forced sequentially thru the one queue. HP recommends at least 3 queues (3 LUNs). By having more than 1 queue, multiple I/Os can be sent to the array, allowing seek overlap (head movement in parallel). Having multiple paths (2 controllers) further allows data transfer overlap.

bv
"The lyf so short, the craft so long to lerne." - Chaucer
Betty Fessler
Frequent Advisor

Re: Oracle on an HP AutoRAID Model 12H, HP-UX 11i

Great info, thanks. Now how do I get the points to stick
do I submit after entering the
points?