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

12H Poor Performance on new LUN

SOLVED
Go to solution
Ted Ellis_2
Honored Contributor

12H Poor Performance on new LUN

we added 4 - 36 GB drives to our 12H (max'd it out with this) and then created a new LUN on the unit. The performance on logical volumes using this LUN is aweful. After creating the LUN, here is the basic disk stats (it is dual controller):

--- Disk space usage --------------------
Total physical = 416795 MB *
Allocated to LUNs = 198400 MB *
Used as Active Hot spare = 34732 MB *
Used by non-included disks = 0 MB *
Used for Redundancy = 62820 MB *
Unallocated (avail for LUNs) = 120843 MB *
-----------------------------------------

Anyone have any comments? I am wondering if we crossed the line and allocated too much space to LUNs?

Ted
13 REPLIES
Ted Ellis_2
Honored Contributor

Re: 12H Poor Performance on new LUN

here is the arraydsp for this system:

Vendor ID = HP
Product ID = C5447A
Array serial number = 0000000E04B6
Array alias = TRIBASE_12H
----------------------------------------------------
Array State = READY
Server name = sfohp9
Array type = 3
Mfg. Product Code = IJMTU00004

--- Disk space usage --------------------
Total physical = 416795 MB *
Allocated to LUNs = 198400 MB *
Used as Active Hot spare = 34732 MB *
Used by non-included disks = 0 MB *
Used for Redundancy = 62820 MB *
Unallocated (avail for LUNs) = 120843 MB *
-----------------------------------------


Vendor ID = HP
Product ID = C5447A
Array serial number = 0000000E04B6
Array alias = TRIBASE_12H
----------------------------------------------------

LUN 1:
Capacity (MB=2^20) = 48000 MB *
Maximum block addr = 98304000
Block length = 512

LUN 2:
Capacity (MB=2^20) = 48000 MB *
Maximum block addr = 98304000
Block length = 512

LUN 3:
Capacity (MB=2^20) = 102400 MB *
Maximum block addr = 209715200
Block length = 512

Total capacity allocated to LUNs = 198400 MB *


Information for disk in slot A1:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST136403LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = 979600001001
Firmware revision = HP01
Drive ID number = AA83 1
Volume set serial no. = AA83 3

Information for disk in slot B1:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST136403LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = 887900001027
Firmware revision = HP01
Drive ID number = AA83 2
Volume set serial no. = AA83 3

Information for disk in slot A2:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST136403LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = 885400001027
Firmware revision = HP01
Drive ID number = AA83 4
Volume set serial no. = AA83 3

Information for disk in slot B2:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST136403LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = 689000007027
Firmware revision = HP01
Drive ID number = AA83 3
Volume set serial no. = AA83 3

Information for disk in slot A3:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST336704LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = J8B500007110
Firmware revision = HP01
Drive ID number = FFFFFFFD18000003
Volume set serial no. = AA83 3

Information for disk in slot B3:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST336704LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = JEKR00007111
Firmware revision = HP01
Drive ID number = FFFFFFFD18000003
Volume set serial no. = AA83 3

Information for disk in slot A4:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST336704LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = JDKK00007111
Firmware revision = HP01
Drive ID number = FFFFFFFD18000003
Volume set serial no. = AA83 3
Recovery Maps are on this disk.

Information for disk in slot B4:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST336704LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = J4HR00007111
Firmware revision = HP01
Drive ID number = FFFFFFFD18000003
Volume set serial no. = AA83 3
Recovery Maps are on this disk.

Information for disk in slot A5:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST336704LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = GWJG00007111
Firmware revision = HP01
Drive ID number = FFFFFFFD18000003
Volume set serial no. = AA83 3

Information for disk in slot B5:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST336704LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = J58K00007111
Firmware revision = HP01
Drive ID number = FFFFFFFD18000003
Volume set serial no. = AA83 3

Information for disk in slot A6:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST136403LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = 05590000W027
Firmware revision = HP01
Drive ID number = FFFFFFFD17A93CF0
Volume set serial no. = AA83 3

Information for disk in slot B6:
Disk State = INCLUDED
Vendor = SEAGATE
Product ID = ST336704LC
Capacity = 34733 MB *
Max. block addr = 71132959
Block length = 512
Serial number = H38K00007111
Firmware revision = HP01
Drive ID number = FFFFFFFD18000003
Volume set serial no. = AA83 3

Total capacity of all installed physical disks = 416795 MB *


Information for controller X:
Vendor ID = HP
Product ID = C5447A
Array serial number = 0000000E04B6
Alias = TRIBASE_12H
Primary = TRUE
SCSI Address = 1
Serial number = R09UADK436511111
Paired serial number = R09UADK438561111
Product revision = HP62


Information for controller Y:
Vendor ID = HP
Product ID = C5447A
Array serial number = 0000000E04B6
Alias = TRIBASE_12H
Primary = FALSE
SCSI Address = 0
Serial number = R09UADK438561111
Paired serial number = R09UADK436511111
Product revision = HP62


Vendor ID = HP
Product ID = C5447A
Array serial number = 0000000E04B6
Array alias = TRIBASE_12H
----------------------------------------------------
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 = DISABLED
WDTR = DISABLED
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 = Normal
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 = 0
Disable Read Hits = FALSE
Resiliency Threshold = 4

Vendor ID = HP
Product ID = C5447A
Array serial number = 0000000E04B6
Array alias = TRIBASE_12H
----------------------------------------------------
Raid 0/1 blocks = 205421184
Raid 0/1 block length = 512
Raid 0/1 capacity = 100303 MB *

Vendor ID = HP
Product ID = C5447A
Array serial number = 0000000E04B6
Array alias = TRIBASE_12H
----------------------------------------------------
Fan F1 = GOOD
Fan F2 = GOOD
Fan F3 = GOOD
Power supply PS1 = GOOD
Power supply PS2 = GOOD
Power supply PS3 = GOOD

Controller X:
Overall state = GOOD
Battery #0 state = GOOD
Battery #1 state = GOOD
DRAM #0 state = GOOD
NVRAM #0 state = GOOD
NVRAM #1 state = GOOD
Controller Y:
Overall state = GOOD
Battery #0 state = GOOD
Battery #1 state = GOOD
DRAM #0 state = GOOD
NVRAM #0 state = GOOD
NVRAM #1 state = GOOD
Joshua Scott
Honored Contributor

Re: 12H Poor Performance on new LUN

You've definitly crossed the line. Your "allocated to luns" must be smaller than the "used for redundancy" + "Unallocated" to acheive RAID 1/0 all the time. What size was that last LUN that you created?

Are you dual-attached?

If you desparately need more performance, you can turn off the Active Hot Spare, which would free up more space, and consequently put more data in RAID 1/0. You could then add 10GB more to the luns and still acheive raid 1/0.

Josh
What are the chances...
Ted Ellis_2
Honored Contributor

Re: 12H Poor Performance on new LUN

thanks for the reply... the 12H documentation indicates that the hot spare space is still used for RAID 0/1 until needed... but I will turn it off and see if we recover anything.
Joshua Scott
Honored Contributor

Re: 12H Poor Performance on new LUN

Ted,

oops, i forgot about that. nevermind on the hotspare.

What are the chances...
Ted Ellis_2
Honored Contributor

Re: 12H Poor Performance on new LUN

:-)

this is dual controller with dedicated SCSI adapter from server to each controller. The last LUN built is shown in the arraydsp details....it is 100 GB
Joshua Scott
Honored Contributor

Re: 12H Poor Performance on new LUN

I have heard that adding new drives to the 12H and then creating luns sometimes doesn't spread the LUNs around to all the disks effectively.

It will rebalance itself as you go along.
What are the chances...
Joshua Scott
Honored Contributor

Re: 12H Poor Performance on new LUN

it seems our replies are out of sync with each other.

How are you doing load-balacing between the controllers?
What are the chances...
Ted Ellis_2
Honored Contributor

Re: 12H Poor Performance on new LUN

that is in line with what I am thinking and seeing. As a test, I have released the hot spare and will let the system work that way for a while. If things return to normal, I will re-enable the hot spare and see if it holds that this is not a performance hit.

cheers
Ted Ellis_2
Honored Contributor

Re: 12H Poor Performance on new LUN

since this is the older 12H Autoraid and not a VA, I can not run this as an Active-Active system... one controller acts as primary and the other as standbye... at least that is how I read it....Controller X is primary and Y is not
Joshua Scott
Honored Contributor

Re: 12H Poor Performance on new LUN

This guy had a similar problem:

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=9759

on the other hand, Perhaps you should break the newest lun into 2 luns and set them up so they are accessed through different HBAs.

I could just be a bandwidth problem.
What are the chances...
Ted Ellis_2
Honored Contributor

Re: 12H Poor Performance on new LUN

on the server side, I have one LUN I/O going across one adapter and the other 2 going across the other, so there is some I/O balance from the host. I am confident that is not an issue for this....
A. Clay Stephenson
Acclaimed Contributor
Solution

Re: 12H Poor Performance on new LUN

You crossed the threshhold for full RAID 1/0 (although not by a huge amount) but a 2nd problem is that you created 1 LUN rather than 2. I assume that this is HP-UX. You should have created 2 LUN's of equal size. LUNa's primary SCSI path should be thru Controller X (alternate Y) and LUNb's should be exactly the opposite. You then add both of these to a Volume Group and stripe all LVOL's across both LUN's in 64K chunks. When you do this, both external SCSI paths are fully utilized as well as all four internal SCSI buses.
If it ain't broke, I can fix that.
Ted Ellis_2
Honored Contributor

Re: 12H Poor Performance on new LUN

thanks Clay.. I found a similar thread where you covered the same ground... agree, that was not wise on our part. If we get a chance we will drop the LUN and rebuild it per recommendation.... thanks