System Administration
cancel
Showing results for 
Search instead for 
Did you mean: 

Can it be that our swap and dump device have not optimum sizes

Frank de Vries
Respected Contributor

Can it be that our swap and dump device have not optimum sizes

We had a message: Deferred swap reservation failure pid;
I looked this up and I was adviced to analyze our swapinfo

At first glance I think our swap lvol2=1gb (too small) and dump lvol10=8gb (too large)

Is it a good idea to change that around so that swap is on lvol10 and dump on lvol2

I would like to add secondary swap but the SAN space is pre-planned and allocated. Would like to avoid budget spend for this year. So above exampl. would give me good fix for a while, right.

see output of our swapinfo -tam and lvlnboot -v
bxud0022:/root#swapinfo -tam
Mb Mb Mb PCT START/ Mb
TYPE AVAIL USED FREE USED LIMIT RESERVE PRI NAME
dev 1024 254 770 25% 0 - 1 /dev/vg00/lvol2
dev 8192 502 7690 6% 0 - 1 /dev/vg00/lvol10
reserve - 2138 -2138
memory 8161 3366 4795 41%
total 17377 6260 11117 36% - 0 -


bxud0022:/root#lvlnboot -v
Boot Definitions for Volume Group /dev/vg00:
Physical Volumes belonging in Root Volume Group:
/dev/dsk/c3t0d0s2 (0/4/1/0.0.0.0.0) -- Boot Disk
Boot: lvol1 on: /dev/dsk/c3t0d0s2
Root: lvol3 on: /dev/dsk/c3t0d0s2
Swap: lvol2 on: /dev/dsk/c3t0d0s2
Dump: lvol10 on: /dev/dsk/c3t0d0s2, 0

thx
Look before you leap
5 REPLIES
Pete Randall
Outstanding Contributor

Re: Can it be that our swap and dump device have not optimum sizes

That won't work because /stand, swap and / must be contiguous.


Pete

Pete
Frank de Vries
Respected Contributor

Re: Can it be that our swap and dump device have not optimum sizes

Yes right, I recall now.
I am confused by the apparent conflicting output of swapinfo -tam and lvlnboot -v and entry in /etc/fstab

In fstab it suggests that lvol10 is my swap, while lvlnboot -v says it is dump. Is there a dual usage , so it is swap anyway ?

How can I measure if I need secondary swap and how much I should need for sec. swap?

# See fstab(4) and sam(1M) for further details on configuring devices.
/dev/vg00/lvol3 / vxfs delaylog 0 1
/dev/vg00/lvol1 /stand vxfs tranflush 0 1
/dev/vg00/lvol10 ... swap defaults 0 0
/dev/vg00/lvol4 /opt vxfs delaylog 0 2
/dev/vg00/lvol5 /tmp vxfs delaylog 0 2
/dev/vg00/lvol6 /usr vxfs delaylog 0 2
/dev/vg00/lvol7 /var vxfs delaylog 0 2
/dev/vg00/lvol8 /home vxfs delaylog 0 2
/dev/vg00/lvol9 /var/adm/crash vxfs delaylog 0 2

thx
Look before you leap
Frank de Vries
Respected Contributor

Re: Can it be that our swap and dump device have not optimum sizes

This link provided all the answers
(by the way a lot of bull is being spread about memory and swap)

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1344865
Look before you leap
Pete Randall
Outstanding Contributor

Re: Can it be that our swap and dump device have not optimum sizes

Your lvol10 is a secondary swap space that can be used for dump, should the need arise - like when you crash. Your present swap setup doesn't seem all that over-utilized (25% and 6%). Why do you think you might need more?

I would be more interested in why it's being utilized at all. Try running vmstat 1 20 and taking a look at the po column. A high po value indicates a log of page-outs and that indicates a lack of memory, not a need for more swap.


Pete

Pete
chris huys_4
Honored Contributor

Re: Can it be that our swap and dump device have not optimum sizes

Hi Frank,

This is the better itrc forum reference. ;)

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1436201

And your system is actually "almost good" configured in a "swap/dump kind of way".

The thread above will also answer why lvol10 should also be the largest one.

So yes, lvol10 will also be used for swap reasons, the entry in lvlnboot -v, just means that when a crashdump needs to be saved, the system will be using lvol10 instead of "standard" lvol2.

For the rest, the lvol10 is actually a bit to small for saving a full crashdump, you need total physical memory, 10gb in this case, + 1Gbyte to be sure, so a 11gb lvol10 instead of the current 8gb lvol10.

And adding a extra 3gbyte to lvol10, could also satisfy the extra swap reservation demands the applications have.

Greetz,
Chris