hp-ux 11.31 dump device using San external LUN

[likid0]

Thnx Dennis, I thought that was the case with 11.31 but wasn't sure.

[chris huys_4]

> Thnx Dennis, I thought that was the case
> with 11.31 but wasn't sure.
It isnt its just dennis opinion. (and im offcourse of the opposite opinion) ;)

[jose da silva]

Thanks All for your inputs and guidelines.

Chris you stated the following which is very important:-
NOTE: In HP-UX 11.31, it might even be that the dumpvolumes dont need to part anymore of /dev/vg00, well with vxvm this is the case, didnt test it with LVM. The dumpluns, where the dumpvolumes are created upon, however should have been made "lif'able" "upfront", through f.e. pvcreate -B /dev/rdisk/disk114;mkboot -e -l /dev/rdisk/disk114, to have them be part of the "persistent dump device list" (crashconf -v) , which is a absolute necessity, to have the dumpconfiguration "survive" reboots.

Some of the docs stated the following HP-UX 11.31:

1. Select an unused disk example /dev/disk/disk52
2. pvcreate /dev/rdisk/disk52
3. vgcreate -p 32 -s 65535 /dev/vgdump
4. lvcreate -C y -r n -n dump_lvol /dev/vgdump
5. Add the new dump lvols in /etc/fstab as shown below:
( ie: /dev/vgdump/dump_lvol... dump defaults 0 0 )
6.crashconf â a ; (add to dump configuration)
7.crashconf â s ; crashconf â v (For persistent dump across reboots)

Is the above correct, I had a look at some of the docs and it is not mentioned to pvcreate -B and mkboot cmds.Please advise

[chris huys_4]

Hi Jose,

It doesnt look like its possible for the moment, with LVM, to have "persistent dump devices" "outside" of /dev/vg00.

NOTE: The "persistency" of the "dump devices", to have the dumpconfiguration survive reboots, is, imo, a "necessity" for dumpdevices outside of /dev/vg00..

From the latest HP-UX 11.31 crashconf patch, which does a update of the crashconf manpage.

PHCO_38958 11.31 crashconf(1M) cumulative patch
[..]
Defect
( QX:QXCR1000847727 )
crashconf (1M) man page does not contain the information
"The logical volumes which are not part of the root volume
group cannot be configured as persistent dump devices".
Resolution:
crashconf (1M) man page is updated with the below
information
"The logical volumes which are not part of the root volume
group cannot be configured as persistent dump devices".

You could open a call with HP support and do a crosscheck and if the above is all "true", ask for the "LVM" feature to be added in a future patch.

For the rest, well you could always join the dark side, i.e. switch from lvm to vxvm as bootdisks and thus also for the dumpdevices, were it does work outside "rootdg" . ;)

on vxvm.

ronin # crashconf -v
Crash dump configuration has been changed since boot.

CLASS PAGES INCLUDED IN DUMP DESCRIPTION
-------- ---------- ---------------- -------------------------------------
UNUSED 1638412 no, by default unused pages
USERPG 145921 no, by default user process pages
BCACHE 19943 no, by default buffer cache pages
KCODE 11204 no, by default kernel code pages
USTACK 1004 yes, by default user process stacks
FSDATA 14 yes, by default file system metadata
KDDATA 262472 yes, by default kernel dynamic data
KSDATA 7955 yes, by default kernel static data
SUPERPG 4744 no, by default unused kernel super pages

Total pages on system: 2091669
Total pages included in dump: 271445

Dump compressed: ON

Dump Parallel: ON

DEVICE OFFSET(kB) SIZE (kB) LOGICAL VOL. NAME
------------ ------------ ------------ ------------ -------------------------
1:0x00000c 2350176 4194304 2:0x000001 /dev/vx/dsk/rootdg/swapvol
1:0x000048 515168 8388608 2:0x2892b8 /dev/vx/dsk/secdumpswapdg/secdumpswapvol
------------
12582912
Persistent dump device list:
/dev/vx/dsk/secdumpswapdg/secdumpswapvol


Greetz,
Chris

[VK2COT · Accepted solution]

Hello,

It is actually an excellent idea to separate dump space from boot volume group.

I wrote two presentations on this topics
recently.

Note that compressed dump has ONE VERY
CRITICAL REQUIREMENT:

Five CPUs per each dump unit

My presentation on swap/dump design is enclosed herewith.

I hope you find it useful.

I also wrote about my view and method for
best practices for HP-UX server design and
build. I can send it to this forum if
there is any interest.

VK2COT

[jose da silva]

Hello V2COT

Thankyou for the valuable information, the slide presentations are execellent.

Can you also share with us your doc with your views and methods for best practices for the HP-UX server design and build.

In one of my previous responces I stated that the HP-UX 11v31 Admin Guide does not contain full examples on how to create the swap and dump spaces from scratch, they only refer to the lvcreate, crashconf and savecrash cmds it would be helpfull if it contained examples.

Warm Regards,
Jose

[VK2COT]

Hello,

I am glad that you find the swap design document useful.

Here is the presentation on HP-UX build
and good practices. It contains some of
the things I deal with when I design and
build HP-UX servers.

To add additional dump areas, the steps are
quite simple. An example (I am not at my desk so the commands might be missing some flags):

# lvcreate -C y -r n -L 8192 -n lvdump2 /dev/vgextP01

# vi /etc/fstab
/dev/vgextP01/lvdump2 ... dump defaults 0 0

# crashconf -a

Cheers,

VK2COT

[jose da silva]

Hello All Concerned

Thanks for the information that I received which was helpfull. A special thanks to VK2COT for the slide presentations which gave us all a better understanding on the swap and dump settings for HP-UX 11v31.

Warm Regards to ALL
Jose da Silva

[jose da silva]

Thanks Once again to all involved. As per the information I have received I have configured the swap and dump spaces.

Warm Regards,
Jose