Thu, 04 Sep 2008

Latest Tunable recommendations for AIX v6 and AIX v5.3

BOTH

NETWORK
no   -p -o rfc1323=1
no   -p -o sb_max=1310720
no   -p -o tcp_sendspace=262144
no   -p -o tcp_recvspace=262144
no   -p -o udp_sendspace=65536
no   -p -o udp_recvspace=655360

Also check the actual NIC interfaces and make sure they are set to at least these values

For AIX v5.3			
No need to set memory_affinity=0 after 5.3 tl05
MEMORY
vmo -p -o minperm%=3
vmo -p -o maxperm%=90
vmo -p -o maxclient%=90
vmo -p -o minfree=960
vmo -p -o maxfree=1088
vmo -p -o lru_file_repage=0
vmo -p -o lru_poll_interval=10
The parameters below should be reviewed and changed (see vmstat –v and lvmo –a later)
PBUFS
ioo -p -o pv_min_pbuf=1024
JFS2
ioo -p -o j2_maxPageReadAhead=128
j2_dynamicBufferPreallocation=16
	Default that may need tuning
	Replaces tuning j2_nBufferPerPagerDevice
JFS
ioo -p -o numfsbufs=1024
ioo -p -o maxpgahead=16

For AIX v6
Make the network changes above
Memory defaults are already correctly set and should not be changed
If you upgrade from a previous version of AIX using migration then you need to check the settings though
The parameters below should be reviewed and changed (see vmstat –v and lvmo –a later)
PBUFS
Tune these using lvmo for the individual volume group 
pv_min_pbuf is now a restricted tunable
JFS2
ioo -p -o j2_maxPageReadAhead=128
	(default above may need to be changed for sequential)
j2_dynamicBufferPreallocation=16
	Default that may need tuning
	Replaces tuning j2_nBufferPerPagerDevice 
JFS
ioo -p -o numfsbufs=1024			(now restricted)
ioo -p -o maxpgahead=16			(now restricted)


CHANGES TO CONSIDER TO ADAPTERS AND HDISKS
Check to make sure your vendor supports these:

fcs0 etc adapters
init_link     pt2pt      INIT Link flags                                    True
max_xfer_size 0x200000   Maximum Transfer Size                              True
num_cmd_elems 2048       Maximum number of COMMANDS to queue to the adapter True

Be careful with init_link as it defaults to AL - this setting is switch and SVC dependent.

Check queue_depth for hdisks - lsattr -El hdisk?

Use iostat -aD to look for sqfull conditions that would indicate the need for an increase in queue_depth