VMFS File Sharing Limits
In previous versions of vSphere, the maximum number of hosts which could share a read-only file on a VMFS volume was 8. The primary use case for multiple hosts sharing read-only files is of course linked clones, where linked clones located on separate hosts all shared the same base disk image. In vSphere 5.1, with the introduction of a new locking mechanism, the number of hosts which can share a read-only file on a VMFS volume has been increased to 32. This makes VMFS as scalable as NFS for VDI deployments & vCloud Director deployments which use linked clones.
Space Efficient Sparse Virtual Disks
A new Space Efficient Sparse Virtual Disk aims to address certain limitations with Virtual Disks. The first of these is the ability to reclaim stale or stranded data in the Guest OS filesystem/database. SE Sparse Disks introduces an automated mechanism for reclaiming stranded space. The other feature is a dynamic block allocation unit size. SE Sparse disks have a new configurable block allocation size which can be tuned to the recommendations of the storage arrays vendor, or indeed the applications running inside of the Guest OS. VMware View is the only product that will use the new SE Sparse Disk in vSphere 5.1.
vSphere Storage APIs – Array Integration
vSphere 5.0 introduced the offloading of snapshots to the storage array for VMware View via the VAAI NAS primitive ‘Fast File Clone’. vSphere 5.1 will allow VAAI NAS based snapshots to be used for vCloud Director in addition to being used for VMware View, enabling the use of hardware/native snapshots for linked clones.
5 Node MSCS Cluster
Historically, VMware only ever supported 2 Node MSCS Clusters. With vSphere 5.1, we are extending this to 5 nodes.
All Paths Down Enhancements
In vSphere 5.1, the objective is to handle the next set of APD use cases involving more complex transient APD conditions. This involves timing out I/O on devices that enter into an APD state. When the timer expires, any I/O sent to the device will be immediately ‘fast failed’ meaning that we do not tie up hostd waiting for I/O. Another enhancement is introducing PDL for some of those iSCSI arrays which present one LUN per target. This was problematic in the past since an APD removed the target as well as the LUN. We are now addressing this scenario.
Storage Protocol Enhancements
FCoE: The Boot from Software FCoE feature is very similar to Boot from Software iSCSI feature which VMware introduced in ESXi 4.1. It allows an ESXi 5.1 host to boot from an FCoE LUN using a NIC with special FCoE offload capabilities and VMware’s software FCoE driver.
iSCSI: We are adding jumbo frame support for all iSCSI adapters in vSphere 5.1, complete with UI support.
Fibre Channel: VMware introduced support for 16Gb FC HBA with vSphere 5.0. However the 16Gb HBA had to be set to work at 8GB. vSphere 5.1 introduces support for 16GB FC HBAs running at 16Gb.
Advanced IO Device Management (IODM) & SSD Monitoring
IODM introduces new esxcli commands to help administrators troubleshoot issues with I/O devices and fabric. This covers Fibre Channel, FCoE, iSCSI, SAS Protocol Statistics and SMART attributes. For SSD monitoring, a new smartd module in ESXi 5.1 will be used to provide Wear Leveling and other SMART details for SAS and SATA SSD. Disk vendors also have the ability to install their own SSD plugins to display vendor specific SSD info.
Storage I/O Control Enhancements
The latency thresholds for the SIOC can now be automatically set. The benefit is that SIOC now figures out the best latency threshold for a datastore as opposed to using a default/user selection for latency threshold. SIOC is now also turned on in ‘stats only mode’ automatically. It doesn’t enforce throttling but does gather more granular statistics about the datastore. Storage DRS can leverage this as it will now have statistics in advance of a datastore being added to a datastore cluster.
Storage DRS Enhancements
vCloud Director will use Storage DRS for the initial placement of linked clones during Fast Provisioning & for managing space utilization and I/O load balancing. Storage DRS also introduces a new datastore correlation detector which means that if a source and destination datastores are backed by the same physical spindles, Storage DRS won’t consider it for migration. Storage DRS also has a new metric (VMobservedLatency) for I/O latency which will be used for more granular I/O load balancing.
Storage vMotion Enhancements
In vSphere 5.1 Storage vMotion performs up to 4 parallel disk migrations per Storage vMotion operation.
That completes the list of storage enhancements in vSphere 5.1. Obviously this is only a brief overview of each of the new features. I will be elaborating on all of these new features over the coming weeks and months.