====== iSCSI as PV for KVM Host LVM ======

**Direct connetion as pool**

<WRAP prewrap center>
<code xml>
<pool type='iscsi'>
    <name>iscsi1</name>
    <source>
        <host name='192.168.1.10'/>
        <device path='iqn.xxxxx'/>
    </source>
    <target>
        <path>/dev/disk/by-path</path>
    </target>
</pool>
</code>
</WRAP>

**I prefer iSCSI as backing storage for LVM**

  * iSCSI target attached as /dev/sdc on KVM Host
  * partitioning /dev/sdc 
<WRAP prewrap center>
<code>
# fdisk -l /dev/sdc
  ...
   Device Boot      Start         End      Blocks   Id  System
/dev/sdc1            2048     4194303     2096128   8e  Linux LVM
</code>
</WRAP>
  * Virt-manager GUI:
    * Host -> Details -> Storage -> Add pool
      * Step1: Name=iscsi1, Type=logical:LVM Volume Group
      * Step2: Target Path=/dev/iscsi1vg, Source Path=(NO /dev/sdc1, better by-path)'/dev/disk/by-path/iqnxxx', Build Pool: flagged -> Finish
    * Select iscsi1 in left pane then NewVolume: Name=tubu1b
    * VM(tubu1) -> Open -> Details -> Add Hardware -> Storage
      * Allocate entire disk now= flagged
      * Select managed or..
      * Browse to volume tubu1b:  /dev/iscsi1vg/tubu1b
      * Device type= Virtio disk
      * Cache mode= default
      * Storage format: leave blank
      * Finish
  * Check KVM Host 
<WRAP prewrap center>
<code>
# pvdisplay -m
  --- Physical volume ---
  PV Name               /dev/sdc1
  VG Name               iscsi1vg
  PV Size               2.00 GiB / not usable 3.00 MiB
  Allocatable           yes (but full)
  PE Size               4.00 MiB
  Total PE              511
  Free PE               0
  Allocated PE          511
  ...
  --- Physical Segments ---
  Physical extent 0 to 510:
    Logical volume	/dev/iscsi1vg/tubu1b
    Logical extents	0 to 510
  ...
</code>
</WRAP>
  * On Guest tubu1 format and mount new disk
<WRAP prewrap center>
<code>
# fdisk -l /dev/vdb
Disk /dev/vdb: 2143 MB, 2143289344 bytes
...
   Device Boot      Start         End      Blocks   Id  System
/dev/vdb1            2048     4186111     2092032   83  Linux

# mkfs.ext4 /dev/vdb1
# mount /dev/vdb1 /media/test
</code>
</WRAP>
  * Check copy with cluster failover
Trig a cluster failover (crm node standby node1) during copy1, VM freeze for 2 seconds
<WRAP prewrap center>
<code>
# pv ubuntu-12.04-server-amd64.iso > /media/test/ubuntu.iso
</code>
</WRAP>
Trig a second cluster failover (crm node standby node2) during copy2, VM freeze for 2 seconds
<WRAP prewrap center>
<code>
# pv ubuntu-12.04-server-amd64.iso > /media/test/ubuntu1.iso
</code>
</WRAP>
Verify data integrity
<WRAP prewrap center>
<code>
diff /media/test/ubuntu.iso /media/test/ubuntu1.iso
</code>
</WRAP>

**NOTE: a bug http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=663931 then you need /etc/init.d/udev restart on KVM host if virt-manager timeout**

<WRAP right 35% rightalign lo>
[[hypervisor:clone|VM clone]] ⇒
</WRAP>
<WRAP left 35% leftalign lo>
⇐ [[hypervisor:lvmstart|Main LVM configuration]]
</WRAP>
<WRAP center 30% centeralign lo>
⇑ [[hypervisor:start|Hypervisor]]
</WRAP>