Chapter 2. The importance of storage in virtual infrastructures
2.2 Hypervisor data storage capabilities
VMware
Citrix XenServer 5.5
Microsoft Hyper-V R2
2.2.1 VMware
The virtual storage manager in VMware resides in the vCenter management console. The vStorage Virtual Machine File System (VMFS) is a cluster file system that provides optimized virtual machine storage virtualization. vStorage stores the virtual machine state in a central location. Then, storage administrators can run multiple, concurrent instances of the VMware ESX Server and access the same virtual machine storage.
VMware data storage management also includes advanced features such as thin provisioning. Another advanced feature is hot extension for virtual logical unit numbers (LUNs) so that storage managers can expand LUNs at the VMFS level without taking down the LUN. In addition, VMware Data Recovery takes virtual machine snapshots and does file-level recovery from virtual machine disk (VMDK) backups.
2.2.2 Citrix XenServer 5.5
Citrix XenServer 5.5 has limited data storage management capabilities unless you purchase the Citrix Essentials add-on package, which includes the Citrix StorageLink technology. With StorageLink, storage arrays behave as native XenServer storage. This way, administrators can provision virtual machine storage directly from the XenServer management console. In this approach, they make available such features as thin provisioning, data deduplication, and performance optimization that are included with the arrays.
2.2.3 Microsoft Hyper-V R2
Microsoft Hyper-V R2 included improvements in the Hypervisor storage capabilities, which are mostly features that VMware customers are already accustomed to. Data storage managers who use Hyper-V can now dynamically add and remove disks from guest operating systems and expand virtual hard disks. Hyper-V R2 also introduced Cluster Shared Volumes so that multiple virtual machines can share LUNs.
Hyper-V R2 with System Center Virtual Machine Manager 2008 R2 also supports Live Storage Migration. Live Storage Migration allows for LUN-to-LUN migration of storage from one platform to another, which requires some downtime. VMware Storage VMotion supports similar
migration without downtime. On vSphere 4.1, migration across different storage vendors is also supported. Hyper-V also uses the Microsoft System Center Data Protection Manager and Microsoft Volume Shadow Copy Service technology for virtual machine snapshots.
2.3 VMware
This section provides information about the new features, functions, and architecture of VMware.
2.3.1 What is new in VMware vSphere vStorage
VMware vSphere vStorage has several new features.
Virtual disk thin provisioning
With VMware thin provisioning, virtual machines can use storage space on an as-needed basis, further increasing utilization of storage for virtual environments. vCenter Server 4.0 enables alerts and provides alarms and reports that specifically track allocation and current usage of storage capacity. This way administrators can optimize the allocation of storage for virtual environments. With thin provisioning, users can safely optimize available storage space, by using over-allocation, and reduce storage costs for virtual environments.
VMware Paravirtualized SCSI
VMware Paravirtualized SCSI (PVSCSI) adapters are high-performance storage adapters that offer greater throughput and lower processor utilization for virtual machines. These adapters are best suited for environments in which guest applications are I/O intensive. VMware recommends that you create a primary adapter for use with a disk that hosts the system software (boot disk) and a separate PVSCSI adapter for the disk that stores user data, such as a database. The primary adapter is the default for the guest operating system on the virtual machine. For example, for virtual machines with Microsoft Windows 2008 guest operating systems, LSI Logic is the default primary adapter.
VMFS Volume Grow
vCenter Server 4.0 allows dynamic expansion of a Virtual Machine File System (VMFS) volume extent to add capacity to an existing data store. VMFS Volume Grow is a new method for expanding a data store without disrupting currently running virtual machines. After a LUN that backs that data store is expanded through an array management utility, the administrator can use VMFS Volume Grow to expand the VMFS extent on the expanded LUN. The newly available space appears as a larger VMFS volume (data store) along with an associated growth event in vCenter Server systems.
Pluggable Storage Architecture
The Pluggable Storage Architecture (PSA) is an open modular framework that enables third-party storage multipathing solutions for workload balancing and high availability. You can use the vSphere command-line interface (CLI) or vCenter Server to manage paths controlled by the default native multipathing. If array-specific functionality is required, a third-party plug-in using the vStorage API for Multipathing can be configured by using the vSphere CLI.
Hot extend for virtual disks
Hot extend is supported for VMFS flat virtual disks in persistent mode and without any VMFS snapshots. When used with the new VMFS Volume Grow capability, the user has maximum flexibility in managing growing capacity in vSphere 4.0.
Storage stack performance and scalability
The combination of the new in-guest virtualization-optimized SCSI driver and additional ESX kernel-level storage stack optimizations dramatically improves storage I/O performance. It makes even the most I/O-intensive applications, such as databases and messaging applications, prime candidates for virtualization.
Software iSCSI and NFS support with jumbo frames
vSphere 4.0 adds support for jumbo frames with both Network File System (NFS) and iSCSI on 1 Gb and 10 Gb network interface cards (NICs).
Fibre Channel over Ethernet
vSphere 4.0 extends the number of I/O consolidation options available to VMware customers by adding Fibre Channel over Ethernet support on Converged Network Adapters (CNAs). For a list of supported Fibre Channel over Ethernet CNAs with vSphere 4.0, see the VMware website at:
http://www.vmware.com
Managing VMFS volumes with array-based LUN snapshots
The mounting of array-based LUN snapshots (and array-based LUN clones) now occurs easily and in a well-managed way in vSphere 4.0. Such LUNs are now automatically discovered after a storage rescan. Single snapshots (or single clones) can be selected for mounting and use by the ESX host. However, to mount a snapshot (clone), the snapshot must be writable. VMFS must write a new unique identifier, or a new VMFS volume signature, to the snapshot or clone to safely mount it in the same farm as the original LUN. For disaster recovery scenarios, in which the replicated volume is not in the same farm, LUNs can be mounted without writing a new signature.
iSCSI support improvements
Updates to the iSCSI stack offer improvements to both software iSCSI and hardware iSCSI. The iSCSI initiator runs at the ESX layer, and ESX uses a hardware-optimized iSCSI host bus adapter (HBA). The result is a dramatic improvement of the performance and functionality of the software and hardware iSCSI and a significant reduction of processor overhead for software iSCSI.
Increased NFS datastore support
2.3.2 vSphere storage architecture
The VMware vSphere storage architecture consists of layers of abstraction that hide and manage the complexity and differences among physical storage subsystems. Figure 2-1 illustrates this storage architecture.
Figure 2-1 vSphere storage architecture
To the applications and guest operating systems inside each virtual machine, the storage subsystem appears as a virtual SCSI controller connected to one or more virtual SCSI disks. These controllers are the only types of SCSI controllers that a virtual machine can see and access. They include BusLogic Parallel, LSI Logic Parallel, LSI Logic SAS, and VMware Paravirtual.
The virtual SCSI disks are provisioned from datastore elements in the data center. A data store is similar to a storage appliance that delivers storage space for virtual machines across multiple physical hosts.
The
datastore abstraction
is a model that assigns storage space to virtual machines while insulating the guest from the complexity of the underlying physical storage technology. The guest virtual machine is not exposed to Fibre Channel storage area network (SAN), iSCSI SAN, direct-attached storage, and network-attached storage (NAS).Each virtual machine is stored as a set of files in a directory in the data store. The disk storage associated with each virtual guest is a set of files within the directory of the guest. You can operate on the guest disk storage as an ordinary file. You can copy, move, or back up the disk. New virtual disks can be added to a virtual machine without powering it down. In that
case, a virtual disk file (a .vmdk file) is created in VMFS to provide new storagefor the added virtual disk. Alternatively, an existing virtual disk file is associated with a virtual machine. Each data store is a physical VMFS volume on a storage device. NAS data stores are an NFS volume with VMFS characteristics. Data stores can span multiple physical storage
subsystems. A single VMFS volume can contain one or more LUNs from a local SCSI disk array on a physical host, a Fibre Channel SAN disk farm, or iSCSI SAN disk farm. New LUNs added to any of the physical storage subsystems are detected and made available to all existing or new data stores. Storage capacity on a previously created data store can be extended without powering down physical hosts or storage subsystems. If any of the LUNs within a VMFS volume fails or becomes unavailable, only virtual machines that touch that LUN are affected.
An exception is the LUN that has the first extent of the spanned volume. All other virtual machines with virtual disks that reside in other LUNs continue to function as normal.
VMFS is a clustered file system that uses shared storage so that multiple physical hosts can read and write to the same storage simultaneously. VMFS provides on-disk locking to ensure that the same virtual machine is not powered on by multiple servers at the same time. If a physical host fails, the on-disk lock for each virtual machine is released so that virtual machines can be restarted on other physical hosts.
VMFS also features failure consistency and recovery mechanisms, such as distributed journaling, a failure consistent virtual machine I/O path, and machine state snapshots. These mechanisms can aid quick identification of the cause and recovery from virtual machine, physical host, and storage subsystem failures.
VMFS also supports raw device mapping (RDM), which is illustrated in Figure 2-2.
RDM provides a mechanism for a virtual machine to have direct access to a LUN on the physical storage subsystem (Fibre Channel or iSCSI only). RDM is useful for supporting two typical types of applications:
SAN snapshot or other layered applications that run in virtual machines
RDM enables better scalable backup by offloading systems using features inherent to the SAN.
Microsoft Clustering Services (MSCS) spanning physical hosts and using virtual-to-virtual and physical-to-virtual clusters
Cluster data and quorum disks must be configured as RDMs rather than files on a shared VMFS.
An RDM is a symbolic link from a VMFS volume to a raw LUN. The mapping makes LUNs appear as files in a VMFS volume. The mapping file, not the raw LUN, is referenced in the virtual machine configuration.
When a LUN is opened for access, the mapping file is read to obtain the reference to the raw LUN. Thereafter, reads and writes go directly to the raw LUN rather than going through the mapping file.
The person with design authority must also make the right decision about the type of storage device and how the storage devices will be selected, configured, and implemented. A wrong decision can damage a virtual infrastructure project. For example, not having the proper knowledge and understanding of the importance of storage can result in performance or availability issues and a loss of money.