Kernel Paramater reccomendations upon performance stats

[Gousset]

Please find attached outputs from running "system.perf.sh" and the current kernel paramaters on a HP-UX 11.11 PA-RISC 4 CPU server with 8Gb of Memory. Could someone reccommend any changes. This server runs an oracle 9.2.0.7 client and java program.

current params are:

======================
Parameter Current Dyn Planned Module
Version
========================================================================
=======
NSTRBLKSCHED - - 2
NSTREVENT 50 - 50
NSTRPUSH 16 - 16
NSTRSCHED 0 - 0
STRCTLSZ 1024 - 1024
STRMSGSZ 65535 - 65535
acctresume 4 - 4
acctsuspend 2 - 2
aio_listio_max 256 - 256
aio_max_ops 2048 - 2048
aio_physmem_pct 10 - 10
aio_prio_delta_max 20 - 20
allocate_fs_swapmap 0 - 0
alwaysdump 0 - 0
bootspinlocks - - 256
bufcache_hash_locks 128 - 128
bufpages 78600 - 78600
chanq_hash_locks 256 - 256
core_addshmem_read 0 Y 0
core_addshmem_write 0 Y 0
create_fastlinks 0 - 0
dbc_max_pct 0 - 0
dbc_min_pct 0 - 0
default_disk_ir 0 - 0
desfree - - 0
disksort_seconds 0 - 0
dnlc_hash_locks 512 - 512
dontdump 0 - 0
dskless_node - - 0
dst 1 - 1
effective_maxpid - - ((NPROC<=30000)?30000:(NPROC*5/4))
eisa_io_estimate - - 0x300
enable_idds 1 - 1
eqmemsize 15 - 15
executable_stack 0 - 0
fcp_large_config 0 - 0
file_pad - - 10
fs_async 0 - 0
ftable_hash_locks 64 - 64
hdlpreg_hash_locks 128 - 128
hfs_max_ra_blocks 8 - 8
hfs_max_revra_blocks 8 - 8
hfs_ra_per_disk 64 - 64
hfs_revra_per_disk 64 - 64
hp_hfs_mtra_enabled 1 - 1
hpux_aes_override - - 0
initmodmax 50 - 50
io_ports_hash_locks 64 - 64
iomemsize - - 40000
ksi_alloc_max 24000 - (NPROC*8)
ksi_send_max 32 - 32
lotsfree - - 0
max_async_ports 50 - 50
max_fcp_reqs 512 - 512
max_mem_window 0 - 0
max_thread_proc 1024 - 1024
maxdsiz 1073741824 - 1073741824
maxdsiz_64bit 2147483647 - 2147483647
maxfiles 1700 - 1700
maxfiles_lim 2048 Y 2048
maxqueuetime - - 0
maxssiz 134217728 - 134217728
maxssiz_64bit 1073741824 - 1073741824
maxswapchunks 16384 - 16384
maxtsiz 67108864 Y 67108864
maxtsiz_64bit 67108864 Y 67108864
maxuprc 2950 Y (NPROC-50)
maxusers 32 - 32
maxvgs 128 - 128
mesg 1 - 1
minfree - - 0
modstrmax 500 - 500
msgmap 3002 - (NPROC+2)
msgmax 8192 Y 8192
msgmnb 16384 Y 16384
msgmni 3000 - (NPROC)
msgseg 12000 - (NPROC*4)
msgssz 8 - 8
msgtql 3000 - (NPROC)
nbuf 0 - 0
ncallout 5250 - (NKTHREAD)
ncdnode 150 - 150
nclist 612 - (100+16*MAXUSERS)
ncsize 27072 - ((8*NPROC+2048)+VX_NCSIZE)
ndilbuffers 30 - 30
netisr_priority - - -1
netmemmax - - 0
nfile 10000 - 10000
nflocks 2000 - 2000
nhtbl_scale 0 - 0
ninode 2500 - 2500
nkthread 5250 - ((NPROC*7)/4)
nni - - 2
no_lvm_disks 0 - 0
nproc 3000 - 3000
npty 500 - 500
nstrpty 60 - 60
nstrtel 60 - 60
nswapdev 10 - 10
nswapfs 10 - 10
nsysmap 6000 - (2*NPROC)
nsysmap64 6000 - (2*NPROC)
o_sync_is_o_dsync 0 - 0
page_text_to_local - - 0
pfdat_hash_locks 128 - 128
public_shlibs 1 - 1
region_hash_locks 128 - 128
remote_nfs_swap 0 - 0
rtsched_numpri 32 - 32
scroll_lines 100 - 100
scs_table_max_size - - 500
scs_table_size - - 10
scsi_max_qdepth 8 Y 8
scsi_maxphys 1048576 - 1048576
sema 1 - 1
semaem 16384 - 16384
semmap 6002 - ((NPROC*2)+2)
semmni 6000 - (NPROC*2)
semmns 12000 - (NPROC*4)
semmnu 2996 - (NPROC-4)
semmsl 2048 Y 2048
semume 2996 - (NPROC-4)
semvmx 32768 - 32768
sendfile_max 0 - 0
shmem 1 - 1
shmmax 4294967296 Y 4294967296
shmmni 512 - 512
shmseg 120 Y 120
st_ats_enabled 0 - 0
st_fail_overruns 0 - 0
st_large_recs 0 - 0
st_san_safe 0 - 0
streampipes 0 - 0
swapmem_on 1 - 1
swchunk 2048 - 2048
sysv_hash_locks 128 - 128
tcphashsz 0 - 0
timeslice 10 - 10
timezone 420 - 420
unlockable_mem 0 - 0
vas_hash_locks 128 - 128
vnode_cd_hash_locks 128 - 128
vnode_hash_locks 128 - 128
vps_ceiling 64 - 64
vps_chatr_ceiling 1048576 - 1048576
vps_pagesize 4 - 4
vx_fancyra_enable 0 - 0
vx_ncsize 1024 - 1024
vx_ninode 0 - 0
vxfs_max_ra_kbytes 1024 - 1024
vxfs_ra_per_disk 1024 - 1024
=============================================

[Ludovic Derlyn]

hi

If you have jfs online , specify for database mount options mincache=direct

Avwait and avserv will be equal to 20 ms, sometines for c100txdx it's more important

regards

L-DERLYN

[Steven E. Protter]

Shalom,

vx_ninode 0 - 0

Change it to a reasonable number. For a moderatly busy database server I've found between 1000 and 4000 will do the trick.

Also, you might want to look into the layout of your files on the disk that has high wait time. If its a SAN disk it might benefit from raid 1 or raid 10.

For databases that do a lot of write transactions raid 5 causes these bottlenecks.

I love when my own(originated from HP) stats program gives me a chance to have to look at the output.

Quite a useful little toy.

SEP

[Ludovic Derlyn]

hi,

Just a remark , often it's recommedned to set dbc_min =5 and dbc_max =8

Regards

L-DERLYN

NB: where are you localized in french ?

[Gousset]

Thanks guys. Very usefull tips. I am actually located in the UK. My head office is based in South of France !!