Server Virtualization Storage Solution

Challenges Posed by Server Virtualization

  • Challenges Posed by Server Virtualization
  • VMware VVOL Virtual Volume Application
  • VMware VAAI Storage Hardware Acceleration
  • VMware SRM Application Level Disaster Recovery

Challenges Posed by Server Virtualization

In recent years, virtualization has become one of the major construction models for data centers. Compared with traditional IT architecture, virtualization technology creates a more dynamic and flexible data center by simplifying the IT infrastructure, thereby improving business agility, providing users with more efficient deployment and maintenance, and reducing TCO costs. With the widespread use of server virtualization, data centers are hosting increasingly large-scale services, posing challenges to storage performance and reliability. At the same time, the advanced features provided by virtualization also need to be supported by storage in order to obtain a good user experience.

MS series storage supports mainstream virtualization platforms such as VMware, Citrix, Hyper-V, OpenStack, KVM, XEN, etc. With good compatibility, MS series storage is able to realize seamless connection between front-end virtualization and storage advanced features.

VMware VVOL Virtual Volume Application

By combining MS series storage with VVOL, MS series storage can reduce the management unit of storage from the LUN level to the VM level, which makes the storage management more flexible and efficient. At the same time, it can also make full use of the characteristics of the storage itself, such as snapshots, replication and other features, to improve the performance of the storage in the VMware environment.

The combination of VVOL and snapshot technology, where snapshots are no longer taken in VVOL units, brings significant performance improvements. The performance of VVOL-based snapshot rollback can be improved by several times to dozens of times, and at the same time, deleting and freeing the snapshot space with VVOL-based LUNs is also faster than ordinary LUNs. VVOL-based LUNs can recover only one VM without affecting other VMs on that LUN, fine-tuning LUN management down to the VM level.

The combination of VVOL and replication technology can greatly save the link bandwidth demand during disaster recovery backup. When establishing disaster recovery backup, the initial synchronization of data in the primary and backup centers is a huge data migration project. The combination of VVOL-based and replication can reduce the amount of unnecessary virtual machine disaster recovery data, thus reducing link bandwidth requirements. And VVOL-based LUNs can replicate and restore only the required VMs without affecting other VMs on this LUN.

While VVOL refines the granularity of management at the VM level, auto thinning technology refines the granularity of management at the data block level. By sensing this fine-grained capacity space change of virtual machines, Auto Thinning technology can realize more fine-grained dynamic resource expansion and better achieve the purpose of capacity "Auto Thinning".

VMware VAAI Storage Hardware Acceleration

MS series storage fully supports VAAI features. By enabling hardware acceleration at the storage level, it can significantly increase the speed of full replication and block clearing, and optimize the VMware environment.

Full replication and block clear operations consume a lot of ESXi host resources and take up a lot of bandwidth between the host and storage, which seriously affects VMware applications. MStar's MS series storage products provide hardware acceleration interfaces to significantly improve the efficiency of full replication and block clearing. For full replication operation, the data can be copied directly from the logical volume of the parent VM to the logical volume of the newly cloned VM within the storage array, and the whole process does not require the participation of the VM. For the block clear operation, you can wait for the blank sectors that occupy most of the space inside the storage array to be cleared by the storage array itself, and the ESXi host is only responsible for sending management commands without participating in the actual data reading and writing. This makes it possible to significantly increase the efficiency when creating, cloning, and migrating virtual machines.

MacroSAN's hardware acceleration solution for VMware dramatically improves the speed of creating and cloning VMware virtual machines to increase efficiency; reduces the resource consumption of VMware virtual machines to safeguard the applications; simplifies the management of VMware and improves maintainability.

VMware SRM Application Level Disaster Recovery

VMware-based virtualized IT architecture has a more important status and role than in traditional physical IT architecture. For the VMware platform, when a major disaster or accident occurs in the production center and the system does not work normally, the virtual machines can be quickly recovered through the SRM component of the disaster recovery center, thus effectively reducing the downtime of the business system and disaster recovery time. Therefore, storage devices need to provide VMware SRM plug-ins to provide fast business recovery support in disaster recovery centers.

The remote replication technology of MacroSAN combined with SRM can realize the remote disaster recovery of VMware virtual machines. Deploy a set of MacroSAN MS series storage arrays in each of the production center and disaster recovery center, and store the virtual machine files in the storage devices; deploy SRM (Site Recovery Manager) components in VMware VirtualCenter, and establish relevant connections between the virtual machine systems in the production center and the disaster recovery center through SRM components. The SRM (Site Recovery Manager) component is deployed in VMware VirtualCenter. The virtual machine data is backed up to the disaster recovery center through the remote replication function on the storage array. When a major disaster or accident occurs in the production center and the system can't work normally, the virtual machines can be quickly recovered through the SRM component in the disaster recovery center, thus effectively reducing the downtime of the business system and disaster recovery time.