why rp3440 so slow...

[Ratzie]

I have just installed a RP3440.
2x146GB drives (Which I will mirror using LVM)
16GB of RAM
16GB of Swap

Everything I launch takes forever to load.
SAM take minutes to come up.

The Support Plus GoldKP11i took hours to run.
swlist -d @ .... took a long time to confirm.
swinstall is taking a long time to install the updated RAID SA SCSI depot.

I mean, I run any one of these, and I walk away for coffee...

Here is top for the swinstall of the RAID SA depot.
Load averages: 0.50, 0.46, 0.28
138 processes: 126 sleeping, 11 running, 1 zombie
Cpu states:
CPU LOAD USER NICE SYS IDLE BLOCK SWAIT INTR SSYS
0 0.02 0.0% 0.0% 0.0% 100.0% 0.0% 0.0% 0.0% 0.0%
1 0.99 51.7% 0.0% 48.3% 0.0% 0.0% 0.0% 0.0% 0.0%
--- ---- ----- ----- ----- ----- ----- ----- ----- -----
avg 0.50 25.7% 0.0% 24.2% 50.1% 0.0% 0.0% 0.0% 0.0%

Memory: 124212K (42620K) real, 229644K (72876K) virtual, 14750880K free Page# 1
/13

CPU TTY PID USERNAME PRI NI SIZE RES STATE TIME %WCPU %CPU COMMAND
1 ? 9173 root 239 20 12128K 1628K run 3:57 100.39 100.21 swlist
0 ? 37 root 152 20 1568K 1568K run 0:15 0.67 0.66 vxfsd
0 ? 1653 root 152 20 38448K 7396K run 0:00 0.10 0.10 cimserve
1 ? 3691 root 152 20 24820K 4020K run 0:00 0.10 0.10 vxsvc
1 ? 0 root 127 20 32K 0K sleep 0:40 0.08 0.08 swapper

WHAT IS LOAD?

[Ratzie]

I have just reboot due to install and during post I get this:

Processor Speed State CoProcessor State Cache Size
Number State Inst Data
--------- -------- --------------------- ----------------- ------------
0 1000 MHz Active Functional 67108864 67108864
1 1000 MHz Idle Functional 67108864 67108864

Central Bus Speed (in MHz) : 200
Available Memory : 16777216 KB
Good Memory Required : Not initialized. Defaults to 32 MB.


The Good Memory Required concerns me, what does this mean?

[Michael Steele_2]

Hi Ratzie:

Regarding" "...The Good Memory Required concerns me, what does this mean?...".

I believe the ISL (* Initial System Loader *) needs 32 MB out of your 16 GB to start the boot up process. This is a normal message and can be ignored.

This 'swlist' command does concern me:

CPU TTY PID USERNAME PRI NI SIZE RES STATE TIME %WCPU %CPU COMMAND
1 ? 9173 root 239 20 12128K 1628K run 3:57 100.39 100.21 swlist

%CPU time is 100% for a process that I believe should be complete.

Run a 'swverify' and 'check_patches' and check the /var/adm/sw/*.logs going back to see when any work was performed, specific to a 'swremove'. My first inclination is that some O/S files have been removed. But this can be verified with these commands.

[Michael Steele_2]

Hi Again Ratzie:

Well I just saw your other posting about how this is a new install. Since this is a new install I think I would start at the CD that you are using. Contact the response center. Give them the label information and ask if its got problems or what. There might be an incompatibility going on.

Additionally, since this is a new install, what is your procedure and how are you answering quesitons posed by the installation process?

Let's see your entire install procedure.

[Ratzie]

This is all that I have done. Pretty basic.
The snap shot was done when I was installing the depot.

I have even noticed that when logging in, it takes forever after I enter the passwd. Quick reply if I enter it wrong...

Here is a new one with nothing running:

Load averages: 0.00, 0.00, 0.00
121 processes: 111 sleeping, 9 running, 1 zombie
Cpu states:
CPU LOAD USER NICE SYS IDLE BLOCK SWAIT INTR SSYS
0 0.00 0.0% 0.0% 0.0% 100.0% 0.0% 0.0% 0.0% 0.0%
1 0.00 0.0% 0.0% 0.0% 100.0% 0.0% 0.0% 0.0% 0.0%
--- ---- ----- ----- ----- ----- ----- ----- ----- -----
avg 0.00 0.0% 0.0% 0.0% 100.0% 0.0% 0.0% 0.0% 0.0%

Memory: 94688K (21440K) real, 190712K (46892K) virtual, 13128504K free Page# 1/9

CPU TTY PID USERNAME PRI NI SIZE RES STATE TIME %WCPU %CPU COMMAND
0 ? 37 root 152 20 1888K 1888K run 0:07 0.61 0.61 vxfsd
1 ? 0 root 127 20 32K 0K sleep 0:31 0.15 0.15 swapper
1 ? 1682 root 152 20 24820K 4020K run 0:00 0.10 0.10 vxsvc
1 ? 1219 root 152 20 61892K 5032K run 0:01 0.06 0.06 cimprovagt
1 ? 1612 root 154 20 4632K 808K sleep 0:04 0.05 0.05 ia64_corehw
1 ? 1558 root 154 10 4568K 792K sleep 0:02 0.05 0.05 psmctd
0 ? 430 root 152 20 2068K 260K run 0:00 0.04 0.04 syncer
1 ? 1216 root 152 20 39248K 7316K run 0:00 0.04 0.04 cimserver

************************************


Boot system, interupt boot sequence
Switch Fast Boot ON
FB ON
Main Menu: Enter command or menu > SEA
Main Menu: Enter command or menu > BO P0 (choose the CDROM) (ide.0)
Interact with IPL? n

Advanced Options: Disable DHCP
Install HP-UX:
Media with Network enabled
Advanced Installation

NETWORK CONFIGURATION:
jaccare1
10.123.123.20
10.123.123.1
255.255.255.0

Basic:
Root Swap (MB) 16384 MB

System:
Set Time Zone: CST6CDT
Set Root Password: password
Network Services:
DNS:123.123.123.101 123.123.124.101 search mts.mb.ca mtsallstream.com
NIS: no
XNTP: xxx.xxx.xxx.xxx

Advanced Installation:

/dev/vg00/lvol1 500 hfs /stand
/dev/vg00/lvol2 16000 vxfs swap
/dev/vg00/lvol3 5000 vxfs /
/dev/vg00/lvol3 10000 vxfs /usr
/dev/vg00/lvol4 5000 vxfs /tmp
/dev/vg00/lvol5 20000 vxfs /home
/dev/vg00/lvol6 15000 vxfs /opt
/dev/vg00/lvol7 5000 vxfs /var
/dev/vg00/lvol8 16000 vxfs /var/adm/crash (needed to size of memory)

Support Plus GOLDQPK11i from CD
-------------------------------
Insert HP-UX Support Plus CD
# mkdir cdrom
# ioscan -fnC disk
# mount /dev/dsk/c0t0d0 /cdrom

# /usr/sbin/swinstall -s /cdrom/GOLDQPK11i \
> -x patch_match_target=true -x autoreboot=true

RAID SA SCSI Controller
-----------------------
Confirm depot was downloaded correctly:
# swlist -d @ /home/lhradow1/RAID-01_B.11.11.0709_HP-UX_B.11.11_64.depot

Install:
# swinstall -s /home/lhradow1/RAID-01_B.11.11.0709_HP-UX_B.11.11_64.depot
automatic reboot

Latest Patch Bundle GOLDKP11i 0712
-------------------
# gunzip -c hpux__11.11_03060930.tgz |tar -xvf -
# ./create_depot_hpux.11.11

# swverify -d \* @ /home/lhradow1/GOLDQKP11i_B.11.11.0712.475.depot
# swinstall -s /home/lhradow1/GOLDQKP11i_B.11.11.0712.475.depot -x autoreboot=true -x patch_match_target=true

[whiz_kid]

looks like a hardware problem.....may be one of ur root disks responding bad..

[Michael Steele_2]

Re: "...I have even noticed that when logging in, it takes forever after I enter the passwd..."

Add a debug comment to /etc/profile so we can see where the hang up is w/in the login process.

sh -x (* top of the file *)

[Khairy]

hi ratzie,

check your hardware status:

# echo "map sellall information;wait infolog quit ok" | cstm > /tmp/cstm.txt

If possible, post the file.

And telnet seems to take longer time? It would be good to check your dns settings.

And one more, maybe its not related but i do have problem last time with swinstall due to dns settings.

Run the following and post us the output
# hostname
# nslookup
# cat /etc/resolv.conf
# cat /etc/nsswitch.conf
# dmesg

Rgds

[SUDHAKAR_18]

suspect's vmstat, bdf, top, sar, iostat output also reqd..

[Ratzie]

Sorry it has been too long to sort this out, I believe it was a DNS problem.