Differences

This shows you the differences between two versions of the page.

--- linux_wiki:logical_volume_management_lvm [2018/05/30 09:55]
billdozor [Delete Virtual Disk]
+++ linux_wiki:logical_volume_management_lvm [2019/05/25 23:50]
@@ Line 1: / Line 1: @@
-====== Logical Volume Management (LVM) ======
-**General Information**
-LVM provides a way to create virtual disks and containers in order to provide flexibility with parition sizes and extending/shrinking space without shutting a system down.
-**Checklist**
-  * Distro(s): Any
-  * Other: The hard disk being formatted is not in use/blank. These steps will erase all data on the target disk.
-----
-====== Expanding or New ======
-Whether expanding or creating new storage, the summarized steps are:
-  * Add virtual disk to VM
-  * Scan SCSI Bus (if needed)
-  * Format the Disk with a partition
-  * Create physical volume using the new disk's partition
-----
-===== Add/Expand Disk =====
-Add or expand the disk to the system either physically or virtually if a VM.
-----
-==== Add Disk: Scan SCSI Bus ====
-If adding a disk to a virtual machine and the new disk does not show up:
-\\
-**One Liner Method**<code bash>echo "- - -" > /sys/class/scsi_host/$(grep mpt /sys/class/scsi_host/host?/proc_name | grep -o -E '(host[0-9])')/scan</code>
-\\
-**Manual Method**
-  * Find the host bus number<code bash>grep mpt /sys/class/scsi_host/host?/proc_name</code>
-    * Example return value:<code bash>/sys/class/scsi_host/host2/proc_name:mptspi</code>
-    * "host2" is the important value
-  * Used the returned host# to scan the SCSI bus<code bash>echo "- - -" > /sys/class/scsi_host/host#/scan</code>
-    * Example from returned value<code bash>echo "- - -" > /sys/class/scsi_host/host2/scan</code>
-  * The hyphens represent controller,channel,lun, so – - – indicates all controllers, all channels, and all luns should be scanned.
-==== Expand Disk: Re-Scan SCSI Bus ====
-If you have expanded an existing disk, rescan the scsi bus<code bash>echo 1 > /sys/class/scsi_device/<device>/rescan</code>
-Where "<device>" is your disk such as "sda".
-----
-===== Format the disk =====
-/dev/sdb is used in this example.
-Use gdisk in order for GPT instead of MBR. (seriously, don't use fdisk...let MBR die already)
-<code bash>
-gdisk /dev/sdb
-</code>
-==== Create a New Empty GUID Partition Table ====
-Once in gdisk:
-  * Use option "o" to create a new empty GUID Partition table. (GPT)
-  * "w" to write table entries
-==== Create the LVM Partition ====
-  * Go back into gdisk. (gdisk /dev/sdb)
-  * "n" to create a new partition
-    * Choose the defaults for partition number, start, and end blocks to fill the entire disk
-    * Use code "8e00" for Linux LVM
-  * "w" to write changes and exit
-----
-===== Create a Physical Volume =====
-About: Think of physical volumes as your "virtual disks" to use in LVM.
-Create a physical volume using the parition that the LVM is on.
-pvcreate <device1-path> [device2-path] [deviceX-path]
-Example:
-<code bash>
-pvcreate /dev/sdb1
-</code>
-----
-**Continue on to either Expanding Storage or Creating New Storage.**
-----
-====== Expanding Storage ======
-Expanding storage steps.
-===== Expand: Summarized Steps =====
-To expand storage for a logical volume, the summarized steps are:
-Performed above in "Expanding or New":
-  * Add virtual disk to VM
-  * Re-Scan SCSI Bus (if needed)
-  * Format the Disk with a partition
-  * Create physical volume using the new disk's partition
-Remaining Steps:
-  * Extend the volume group using the new physical volume
-  * Extend the logical volume
-  * Verify new LVM layout
-  * Grow the file system
-  * Cleanup (if a pvmove of the logical volume was done to a new disk)
-    * Verify old physical volume is not used
-    * Reduce the volume group by removing the physical volume
-    * Remove the physical volume LVM label
-    * Delete virtual disk from virtual management UI
-----
-===== Extend the Volume Group =====
-  - Add the new physical disk to the existing volume group<code bash>vgextend vglocal /dev/sdb1</code>
-----
-===== Extend Logical Volume =====
-  - Extend the logical volume that needs the space (/home will be used in this example)
-    - **Option 1(Preferred)**: Move the physical extents of the logical volume to the new disk, so all of it resides on the same VMDK, and then use all of that new disk's space.
-      - There may be a performance hit if a logical volume is spanned across multiple VMDKs, so this method is preferred.<code bash>pvmove --name lvhome /dev/sda2 /dev/sdb1
-lvextend --resizefs /dev/vglocal/lvhome /dev/sdb1</code>
-    - Option 2: Give the logical volume all of the space from the newly added physical volume<code bash>lvextend --resizefs --extents +100%PVS /dev/vglocal/lvhome /dev/sdb1</code>
-    - Option 3: Give the logical volume a specific amount of additional free space from a specific physical volume<code bash>lvextend --resizefs --size +10G /dev/vglocal/lvhome /dev/sdb1</code>
-    - Option 4: Give the logical volume all of the free space available to the volume group (potentially across multiple physical volumes)<code bash>lvextend --resizefs --extents +100%FREE /dev/vglocal/lvhome</code>
-==== Extending a Logical Swap Volume ====
-Extending a swap volume is slightly different, as the "--resizefs" flag will not work.
-  - Example: Specify the total size of the logical swap volume<code bash>lvextend --size 8G /dev/vglocal/lvswap /dev/sdb1</code>
-  - Disable all swap<code bash>swapoff -a</code>
-  - Create a new swap area, overwriting the original<code bash>mkswap /dev/vglocal/lvswap</code>
-  - Enable swap<code bash>swapon -a</code>
-  - Verify<code bash>swapon -s</code>
-----
-===== Verify New LVM Layout =====
-  - Verify LVM Allocation
-    - Logical Volume: <code bash>lvs</code>
-    - Volume Group: <code bash>vgs</code>
-    - Physical Volumes: <code bash>pvs</code>
-**Logical Volume Location(s) on Physical Volumes**
-If you would like to see what logical volumes reside where, a nice command to view physical extent to logical volume mappings is:<code bash>pvdisplay -m</code>
-----
-===== Grow File System =====
-**Skip if you used the "--resizefs" flag to lvextend above.**
-  - Grow the file system
-    - Ext2/3/4<code bash>resize2fs /dev/vglocal/lvhome</code>
-    - XFS<code bash>xfs_growfs /dev/vglocal/lvhome</code>
-  - Verify filesystem space<code bash>df -h</code>
-----
-===== Cleanup =====
-If you have just moved all of the data of a logical volume off of an old physical volume, the old physical volume can go away if it is no longer in use.
-Examples below: You have just moved everything off of /dev/sdc1 to a new disk and can now remove the old disk (/dev/sdc).
-==== Verify Old Physical Volume Not Used ====
-<code bash>
-pvs -o +pv_used
-</code>
-==== Reduce Volume Group ====
-vgreduce vgname olddisk
-<code bash>
-vgreduce vglocal /dev/sdc1
-</code>
-==== Remove Physical Volume ====
-pvremove olddisk
-<code bash>
-pvremove /dev/sdc1
-</code>
-==== Delete Virtual Disk ====
-Login to the virtualization user interface and delete the old disk from inventory.
-**Errors During LVM Commands**
-  * If you see errors during LVM commands like this after a device delete from vmware<code bash>root@llsrlscd01 home $ pvs
-  /dev/sdc: read failed after 0 of 4096 at 0: Input/output error
-  /dev/sdc: read failed after 0 of 4096 at 17179803648: Input/output error
-  /dev/sdc: read failed after 0 of 4096 at 17179860992: Input/output error
-  /dev/sdc: read failed after 0 of 4096 at 4096: Input/output error
-  /dev/sdc1: read failed after 0 of 4096 at 17178755072: Input/output error
-  /dev/sdc1: read failed after 0 of 4096 at 17178812416: Input/output error
-  /dev/sdc1: read failed after 0 of 4096 at 0: Input/output error
-  /dev/sdc1: read failed after 0 of 4096 at 4096: Input/output error
-  PV         VG      Fmt  Attr PSize   PFree
-  /dev/sda2  vglocal lvm2 a--   19.50g 3.50g
-  /dev/sdd1  vglocal lvm2 a--  250.00g    0
-  /dev/sde1  vglocal lvm2 a--   40.00g    0</code>
-    * Fix by rebooting the system or<code bash>echo 1 > /sys/block/sdX/device/delete</code>
-      * Where "sdX" is the device, such as 'sdc' in the above example.
-----
-====== Creating New Storage ======
-===== Create New: Summarized Steps =====
-To create new storage for a logical volume, the summarized steps are:
-Performed above in "Expanding or New":
-  * Add virtual disk to VM
-  * Re-Scan SCSI Bus (if needed)
-  * Format the Disk with a partition
-  * Create physical volume using the new disk's partition
-Remaining Steps:
-  * Create a volume group
-  * Create a logical volume
-  * Create a file system on the logical volume
-  * Mount the logical volume
-----
-===== Create a Volume Group =====
-About: Volume groups aggregate physical volumes into a usable pool of disk space.
-vgcreate <vgname> <physical-device>
-Example:
-<code bash>
-vgcreate vgstorage /dev/sdb1
-</code>
-----
-===== Create a Logical Volume =====
-About: Logical volumes use the disk space made available via a volume group.
-lvcreate --size <size> --name <lv-name> <vg-name>
-Example: Create a 100G logical volume
-<code bash>
-lvcreate --size 100G --name lvbackups vgstorage
-</code>
-Example: Create a logical volume that uses 100% of the volume group free extents
-<code bash>
-lvcreate --extents 100%FREE --name lvbackups vgstorage
-</code>
-----
-===== Create the file system on the logical volume =====
-About: So far, the only things being created were pretty much containers. Create the actual file system below.
-mkfs -t <fstype> -L <label> /dev/mapper/<vgname>-<lvname>
-Example:
-<code bash>
-mkfs -t ext4 -L Backups /dev/mapper/vgstorage-lvbackups
-</code>
-----
-===== Mount the logical volume by uuid =====
-==== Find the UUID with blkid ====
-<code bash>
-blkid
-/dev/mapper/vgstorage-lvbackups: UUID="d89744ad-133c-452b-98d8-9480ef18fe77" TYPE="ext4"
-</code>
-==== Add entry to /etc/fstab by UUID ====
-<code bash>
-vim /etc/fstab
-UUID=d89744ad-133c-452b-98d8-9480ef18fe77 /opt/backups ext4 defaults 0 2
-</code>
-==== Mount the fstab entries and check to see if it mounted with df ====
-<code bash>
-mount -a
-df -h
-</code>
------
-====== LVM Renaming ======
-We don't always get a name right the first time. Luckily, LVM allows for easy renaming of logical volumes and volume groups.
-Rename a Volume Group (vgrename vg-oldname vg-newname)
-<code bash>
-vgrename vgbackupmirror vgredusb
-</code>
-Rename a Logical Volume (lvrename vgname lv-oldname lv-newname)
-<code bash>
-lvrename vgredusb lvbackupmirror lvredbackups
-</code>
-----
-====== LVM Snapshots ======
-LVM allows you to "freeze" the metadata of files. This allows for things like rsync to complete successfully even if files it is trying to sync are modified during the operation.
-{TODO}