OnApp

StorPool provides and maintains integration with OnApp, whereas each VM disk becomes a separate volume in the StorPool storage system.

The LVM layer, traditionally used by OnApp on shared block storage (SAN) systems is bypassed, thus avoiding a number of issues including

• QoS - single IO queue for many VMs

• LVM metadata corruption due to use of non-cluster LVM on shared storage

In addition to removing the dependency on LVM, with StorPool’s approach to the integration, the client gets fine grained control over each individual volume, in terms of monitoring, data placement or QoS parameters.

On top of replacing LVM, the integration provides the following additional benefits:

• CPU and RAM resource isolation between virtual machines and system processes in each hypervisor.

• Speed up of backups by direct communication between OnApp backup servers and the shared storage system, bypassing the iSCSI initiator/target typically used by OnApp to perform backups.

• Speed up of VM migration, by moving it to a faster (10GE) network.

• Speed up of data migration between StorPool datastores.

• Speed limits for data migration and backups.

• OS-independent LV locking emulation.

For detailed information about using StorPool with OnApp, see:

• OnApp acceptance tests
• OnApp LVM to LVMSP disk migration
• OnApp XEN to KVM disk migration
• OnApp KVM to KVM with virtio migration procedure
• Known issues and FAQ