LVM

From Edgar BV Wiki
Revision as of 12:52, 2 March 2007 by Red (talk | contribs) (New page: LVM management: vgscan # scans the devices and adds them to lvmtab.d directory <Before fdisking the disk!> pvcreate /dev/hde /dev/hdf /dev/hdg /dev/hdh # initialize a disk or part...)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

LVM management:

vgscan # scans the devices and adds them to lvmtab.d directory

<Before fdisking the disk!>

pvcreate /dev/hde /dev/hdf /dev/hdg /dev/hdh # initialize a disk or partition for use by LVM

NB if creating on /dev/md0 (raid devices) ensure that md_component_detection = 1 in /etc/lvm/lvm.conf

vgcreate tripserv_vol /dev/hde /dev/hdf /dev/hdg /dev/hdh # creates the volume group tripserv_vol

lvcreate -i3 -I4 -L100M -nvolumename tripserv_vol # create the logical volume where 

# -i = Gives the number of stripes.  This is equal to the number of physical volumes to scatter the logical volume.
  1. -I = Gives the number of kilobytes for the granularity of the stripes.
             StripeSize must be 2^n (n = 2 to 9)
  1. -L = size
  1. -n = name of logical volumename
  1. tripserv_vol = volumegroupname

mke2fs -j -Lfslabel /dev/tripserv_vol/volumename #formats the partition for use

  1. -j = journaling filesystem
  2. -L = labelname
  1. for snapshot volumes first unmount the drive you want to snapshot to then use

lvcreate -L50M -s -n volumename /dev/tripserv_vol/journaleddevicename

  1. mount it directly no need to mke2fs it

then in /etc/fstab

for journaled volumes:

/dev/tripserv_vol/code /export/home/code ext3 defaults 0 0

for non-journaled:

/dev/tripserv_vol/sfx /export/home/sfx ext2 defaults 0 0

for snapshots:

/dev/tripserv_vol/codesnap /export/home/codesnap ext3 ro 0 0

mount -a

  1. extending, reducing and removing

to extend:

(only lvm1) unmount /dev/tripserv_vol/volumename e2fsadm -L+15G /dev/tripserv_vol/volumename

will make the volumename 15G larger

OR

unmount /dev/tripserv_vol/volumename lvextend /dev/tripserv_vol/volumename -L+100M (to enlarge by 100MB)

check the drive and resize2fs the drive to make it larger, then fsck to check integrity

e2fsck -fv /dev/tripserv_vol/volumename resize2fs /dev/tripserv_vol/volumename e2fsck -fv /dev/tripserv_vol/volumename

to reduce: umount /dev/tripserv_vol/volumename resize2fs /dev/tripserv_vol/volumename 20971520 (new size of volume in blocks) lvreduce /dev/tripserv_vol/volumename -L-10G e2fsck -fv /dev/tripserv_vol/volumename mount /dev/tripserv_vol/volumename

Note: you can find the proper size in blocks by doing an lvreduce then e2fsck. It will give the block size in the error message, after which you can lvextend it back to where it was, e2fsck it and then resize2fs it with the correct paramaters.

After these operations the computer needs rebooting!

to remove:

lvremove /dev/tripserv_vol/volumename

  1. Snapshots

You make a new snapshot volume name based on the existing logical volume you want to back up

lvcreate -s -L500M -n docusnap /dev/tripserv_vol/docu

  1. If the pv size displayed < the disk size

check with pvscan / pvdisplay and with lvmdiskscan. pvdisplay shows the metadata on the disk, lvmdiskscan looks at what the OS has to say. If size mismatch upgrade to lvm2 from testing (not Sarge, see "installing packages from testing on stable machines using apt.txt") and run pvresize /dev/md0

tripserv:/etc# pvscan pvscan -- reading all physical volumes (this may take a while...) pvscan -- ACTIVE PV "/dev/hdg1" of VG "tripserv_vol" [38.16 GB / 7.93 GB free] pvscan -- ACTIVE PV "/dev/hdh1" of VG "tripserv_vol" [38.16 GB / 8.01 GB free] pvscan -- ACTIVE PV "/dev/hde1" of VG "tripserv_vol" [55.91 GB / 0 free] pvscan -- ACTIVE PV "/dev/hdf1" of VG "tripserv_vol" [55.91 GB / 3.37 GB free] pvscan -- total: 4 [188.16 GB] / in use: 4 [188.16 GB] / in no VG: 0 [0]

tripserv:/etc# fdisk /dev/hde

The number of cylinders for this disk is set to 116336. There is nothing wrong with that, but this is larger than 1024, and could in certain setups cause problems with: 1) software that runs at boot time (e.g., old versions of LILO) 2) booting and partitioning software from other OSs 

  (e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/hde: 16 heads, 63 sectors, 116336 cylinders Units = cylinders of 1008 * 512 bytes 

  Device Boot    Start       End    Blocks   Id  System

/dev/hde1 1 116336 58633312+ 8e Linux LVM

Command (m for help): q

You have new mail in /var/spool/mail/root tripserv:/etc# fdisk /dev/hdb

The number of cylinders for this disk is set to 116336. There is nothing wrong with that, but this is larger than 1024, and could in certain setups cause problems with: 1) software that runs at boot time (e.g., old versions of LILO) 2) booting and partitioning software from other OSs 

  (e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/hdb: 16 heads, 63 sectors, 116336 cylinders Units = cylinders of 1008 * 512 bytes 

  Device Boot    Start       End    Blocks   Id  System

/dev/hdb1 1 116336 58633312+ 8e Linux LVM

Command (m for help): q

tripserv:/etc#

tripserv:/etc# dmesg | grep hde 

   ide2: BM-DMA at 0xa000-0xa007, BIOS settings: hde:pio, hdf:pio

hde: MAXTOR 6L060J3, ATA DISK drive hde: 117266688 sectors (60041 MB) w/1819KiB Cache, CHS=116336/16/63, UDMA(100) 

hde: hde1

tripserv:/etc# dmesg | grep hdb 

   ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:DMA, hdb:DMA

hdb: MAXTOR 6L060J3, ATA DISK drive hdb: 117266688 sectors (60041 MB) w/1819KiB Cache, CHS=116336/16/63, UDMA(33) 

hdb: unknown partition table
hdb: hdb1
hdb: hdb1
Retrieved from "http://wiki.edgarbv.com/index.php?title=LVM&oldid=55"