Good morning,
We have replaced EMC in a large government organisation and slowly building up an environment that the customer is happy with.
They have experienced less than expected performance (when compared with the old clariion) and I am trying to understand why (it could be perception).
A “sanlun lun show –p” shows me multi-pathing as A/A and failover provider as HP PV Links which I believed is only A/P capable (or was).
rdlisla2# sanlun lun show -p
MDSVRFAS01A:/vol/RDLISLA2UTILS/QTRDLISLA2UTILSLUN11/RDLISLA2UTILSLUN11.LUN (LUN 111)
Agile Filename: /dev/rdisk/disk38 VG: /dev/vgutils
200g (214748364800) lun state: GOOD
Filer_CF_State: Cluster Enabled Filer Partner: MDSVRFAS01B
Multipath_Policy: A/A
Multipath_Provider: HP PV Links
-------- --------- -------------------- ------ ------- ------- -------------
host filer /dev/dsk primary partner PVlinks
path path filename host filer filer path failover
state type or Hardware Path HBA port port priority
-------- --------- -------------------- ------ ------- ------- -------------
up primary /dev/dsk/c19t13d7 fcd0 0d 0
up primary /dev/dsk/c16t13d7 fcd1 0h 1
up secondary /dev/dsk/c20t13d7 fcd0 0d 2
up secondary /dev/dsk/c17t13d7 fcd1 0h 3
A vgdisplay shows,
rdlisla2# vgdisplay -v /dev/vgutils
--- Volume groups ---
VG Name /dev/vgutils
VG Write Access read/write
VG Status available
Max LV 255
Cur LV 3
Open LV 3
Max PV 255
Cur PV 2
Act PV 2
Max PE per PV 8000
VGDA 4
PE Size (Mbytes) 32
Total PE 12798
Alloc PE 10974
Free PE 1824
Total PVG 0
Total Spare PVs 0
Total Spare PVs in use 0
VG Version 1.0
VG Max Size 63750g
VG Max Extents 2040000
--- Logical volumes ---
LV Name /dev/vgutils/lvol1
LV Status available/syncd
LV Size (Mbytes) 32000
Current LE 1000
Allocated PE 1000
Used PV 1
LV Name /dev/vgutils/lvol2
LV Status available/syncd
LV Size (Mbytes) 19168
Current LE 599
Allocated PE 599
Used PV 1
LV Name /dev/vgutils/lvol3
LV Status available/syncd
LV Size (Mbytes) 300000
Current LE 9375
Allocated PE 9375
Used PV 2
--- Physical volumes ---
PV Name /dev/dsk/c19t13d6
PV Name /dev/dsk/c16t13d6 Alternate Link
PV Name /dev/dsk/c20t13d6 Alternate Link
PV Name /dev/dsk/c17t13d6 Alternate Link
PV Status available
Total PE 6399
Free PE 0
Autoswitch On
Proactive Polling On
PV Name /dev/dsk/c19t13d7
PV Name /dev/dsk/c16t13d7 Alternate Link
PV Name /dev/dsk/c20t13d7 Alternate Link
PV Name /dev/dsk/c17t13d7 Alternate Link
PV Status available
Total PE 6399
Free PE 1824
Autoswitch On
Proactive Polling On
Could this be causing performance issues (as trying to use the alternate path and would have to time-out the requests that are sent down that path which – from memory – is around 60 seconds ).
If so, any idea how I can change it to be A/P
I look forward to any replies.
cheers
Mike
