lvextend error

[Carol Yu]

# lvextend -L 10240 /dev/vg04/lvol3
lvextend: Not enough free physical extents available.
Logical volume "/dev/vg04/lvol3" could not be extended.
Failure possibly caused by contiguous allocation policy.
Failure possibly caused by strict allocation policy

[Robert-Jan Goossens]

Hi,

Could you give us the output from
# lvdisplay -v /dev/vg04/lvol3

Regards,
Robert-Jan

[Navin Bhat_2]

When you run
lvdisplay -v /dev/vg04/lvol3

what is the display under Allocation?

If you have something like this

Allocation PVG-strict/contiguous <--- this is why

[Mel Burslan · Accepted solution]

looks like your allocation policy is set to strict/contiguous and most probably for a reason at the time of lvol creation.

you can change them back by command :

lvchange -C n -s n /dev/vg04/lvol3

but the -C n directive in this command may impact your performance and -s n directive will affect the redundancy, i.e., mirror copies will not be placed on different physical volumes.

if you are not worried about performance and do not have mirror copies, you can do this without any concern.

[Carol Yu]

Right, I have confirmed it is related to the strict and contiguous allocation policy.

This is a qa place on the server, so I am not too worried about the performance. But I do have a mirror copy for the lv, so how is this solution going to affect mirror copy?

Thanks a lot,
Carol

[Navin Bhat_2]

It is a matter of extent of data redundancy. Do you want mirror setup so that the logical extents are on the different physical volume (strict) or on different physical volume group (pvg-strict) or you want them shared (the no strict policy).

[Mel Burslan]

Unless there is a lot of logical volumes with and without mirrors, you can get by without setting the strict policy off, i.e., run the command as

lvchange -C n /dev/vg04/lvol3

then try extending it. If you are getting the same error message then you have to turn it off with the command

lvchange -s n /dev/vg04/lvol3

at this point you should be able to extend it. but keep in mind that data redundancy for this logical volume has been compromised at this point.

The only solution is re-arranging the logical volumes on the vg04 from scratch with better planning, or add more disks if possible, in pairs of course to this vg04.

HTH.