11.31 MSA1500 et load balancing

[RUET]

Hello guys

I have a problem, with a server connected to a SAN to access some MSA disks .

I use 11.31 and persistant DSFfor the disks I include in LVM.

When I check witch scsimgr the disks, I have 2 paths for each disks, but only one is flagged to be ACTIVE, and the other one is STANDBY .

I assume (verified by sar -H and get_stat of scsimgr), that all IOs are going to only one path.

When I check attribute, with scsimgr -get_attr, leg_mpath_enable is set to true .. and load_bal_policy is set to round_robin ..

Any ideas ?

thanks in advance guys

pat

[Torsten.]

I hope your MSA firmware is already active/active. Otherwise it would not really work at all. If you use SSP, make sure you enable access for all HBAs.

[RUET]

in SSP, acees is set for both HBA. I can see, with ioscan both path to disks ..

What do you mean by active/active ? Is it the only reason that prevent me load balancing ?

[Torsten.]

In the past the msa1500 was active/passive only, 1 controller was active, the other standby only.

Then with a new firmware the array became active/active.

What is your version?

[Steven E. Protter]

Shalom,

There is a setting in the firmware for the MSA-1500 that lets you decide between active/active and active/passive.

Older versions of the MSA-1500 firmware may have defaulted to active/passive.

It is always a good idea when running into performance issues to update the firmware on the MSA-1500 and the HBA if available.

The MSA-1500 is a pretty basic disk array and load balancing may not provide much of a boost. Have you checked to see if the active controller is maxed out on performance?

This could be a configuration issue. Common mistake is to put a heavy write database on raid 5. This makes writes back up in a queue. Data, index, redo(oracle terms), should be on Raid 1 storage.

SEP

[Torsten.]

"There is a setting in the firmware for the MSA-1500 that lets you decide between active/active and active/passive."

More or less. I think SEP is refering to

" PreferredPathMode

The setting that you select for this option determines how the preferred I/O path to a particular logical
drive is set for a redundant array controller that is in an active/active configuration.
Not all controllers support this feature, and controllers in an active/standby configuration disregard this
option.
â ¢ Auto is the default setting for new configurations. In this case, the storage system automatically
selects the I/O path from the redundant controller to the logical drive and dynamically load
balances all paths.
â ¢ Manual enables you to assign the logical drive to a specific redundant controller. If you select this
setting, use the PreferredPath (on page 39) command to specify the path."

[RUET]

Thanks for all this ..
- How to I check status of my MSA controllers. I'm quite sure it is active/active mode.
- Before I was in 11.23, and I was using a sort of "load balancing" playing with LVM PVlinks and LVM distribution ..

[RUET]

I tried using CLI, abd it's written:
Controller Status : OK
Redundancy Stae: Active
Redundancy mode: Active-Active
Preferred Path Mode: Automatic

[Torsten.]

Connect via CLI to the controller and run

> show tech_support

==>
CLI-1> t show tech_support

CLI-1> show version -all
MSA1500 Firmware Revision: 7.00b1500
,,,

This is firmware version 7

[RUET]

firmware version is 7

[Torsten.]

Can you post the "get_info" output of scsimgr?

Example:

# scsimgr get_info -D /dev/rdisk/disk55

[RUET]

scsimgr get_info -D /dev/rdisk/disk18

STATUS INFORMATION FOR LUN : /dev/rdisk/disk18

Generic Status Information

SCSI services internal state = ONLINE
Device type = Direct_Access
EVPD page 0x83 description code = 1
EVPD page 0x83 description association = 0
EVPD page 0x83 description type = 3
World Wide Identifier (WWID) = 0x600508b300936ee0a97deb0154f8002f
Serial number = "PB9840PX3TQ02U"
Vendor id = "HP "
Product id = "MSA VOLUME "
Product revision = "7.00"
Other properties = ""
SPC protocol revision = 2
Open count (includes chr/blk/pass-thru/class) = 1
Raw open count (includes class/pass-thru) = 0
Pass-thru opens = 0
LUN path count = 2
Active LUN paths = 1
Standby LUN paths = 1
Failed LUN paths = 0
Maximum I/O size allowed = 2097152
Preferred I/O size = 2097152
Outstanding I/Os = 0
I/O load balance policy = round_robin
Path fail threshold time period = 0
Transient time period = 120
Tracing buffer size = 1024
LUN Path used when policy is path_lockdown = NA
LUN access type = T10 Asymmetric Active-Active
Asymmetric logical unit access supported = Both implicit and explicit
Asymmetric states supported = ao_sup, an_sup
Preferred paths reported by device = Yes
Preferred LUN paths = 0

Driver esdisk Status Information :

Capacity in number of blocks = 2048002047
Block size in bytes = 512
Number of active IOs = 0
Special properties =
Maximum number of IO retries = 45
IO transfer timeout in secs = 30
FORMAT command timeout in secs = 86400
START UNIT command timeout in secs = 60
Timeout in secs before starting failing IO = 120
IO infinite retries = false
root@ssrv117:.../rootpr #

[RUET]

hello gurus, no anymore ideas ?
pat

[RUET]

NO nore news ..
I'm still trying to make both path in activation, but I still don't have the solution ..

Can it be a 11.31 problem or wrong configuration ?

[RUET]

No news from you guys ..

no solution to my problem ?

My server in these config is sending 80% lof the load activity to one HBA , and nearly nothing to the other ...

That's not well balanced and so performance are not the best I can get ..