VMware announced the details on the next version of Virtual SAN. This version will be VSAN 6.0.
For those not familiar with VSAN: it allows to build distributed storage from server based disks. So a storage network is not required for running virtual machines and using features like vMotion. Just use cheap local disks, have VSAN manage redundancy and you are ready to go. Advantages of VSAN are costs, agility, reduced complexity.
VSAN while being part of the ESXi kernel requires a separate license. At January 2015 VSAN is used by about 1000+ customers since the release 9 months ago.
Interesting is that VMware integrated the Virsto filesystem into VSAN. Virsto was acquired by VMware.
The message of VMware is that VSAN 6.0 is ready for business critical applications like SQL Server and Oracle. So VSAN is now positioned by VMware to support scenario’s like business critical apps, VDI, target for Disaster Recovery and test/dev environments.
VSAN 6.0 can now be deployed in two configurations
- new is the All-Flash configuration where SSD can be used for both cache (cache tier) and persistent storage (capacity tier). Mind the All-Flash configuration requires 10GBps network. Flash is used a a write buffer.
- hybrid where SSD is used for cache and HDD for storage. Supports both 1 Gbps and 10Gbps networks.Flash is used as read cache and write buffer.
VSAN 6 will not be available in EVO: RAIL configurations before the second half of 2015.
Mind All-Flash is an add-on to Virtual SAN with HDD for storage. Customers have to pay 60% extra for All-Flash.
- VMware Virtual SAN is priced at $2,495 per CPU.
- VMware Virtual SAN for Desktop is priced at $50 per user.
- The new All-Flash architecture will be available as on add-on to VMware Virtual SAN 6 and will be priced at $1,495 per CPU and $30 per desktop. VMware vSphere Virtual Volumes will be packaged as a feature in VMware vSphere Standard Edition and above as well as VMware vSphere ROBO editions.
More info here.
VSAN 6.0 is a major new release which will have the following new features
- 100K IOPS per hosts in All-Flash configuration (increased 5 x compared to VSAN 5.5 )
- 40K IOPS per hosts in Hybrid configuration (SSD for cache, HD for persistent storage (increased 2 x)
- scales to 64 nodes (increased from 32 nodes)
- 200 VM’s per hosts (increased from 100)
- VM’s per cluster now 6000. (increased from 3200)
- support for direct attached JBOD (just a bunch of disks). This is important for support of VSAN on blades. For example IBM Flex10. Not each JBOD will be supported. Check the VSAN Compatibility Guide.
- hardware based checksum and encryption
- proactive rebalance. This allows administrator to have control over the process which makes sure each disk of a datastore has the same amount of data stored on it.
- 62 TB max virtual disk file size
- support for layer 3 network
- 9000 components per host (increased from 3000)
- VMDK virtual disks can now be stored on SSD. In the previous version SSD was used as a cache layer only.
- rack awareness to tolerate rack failures. In VSAN 6 you can define faultdomains. Basically you tell which hosts are located in the same rack. VSAN will then make sure the replica of a VM is stored in a different rack. This will not support streteched clusters. You will need a minimum of three fault domains.
- High performance snapshots and clones (based on the Virsto filesystem)
- new diskformat VSAN-FS
- 32 snapshots per VM
- New RVC commands for management and configurations purposes have been added
- Virtual SAN Health Services which allows for monitoring of VSAN subsystems. Tool is embeded in vCenter.
- PowerCLI 6.0 delivers a set of Virtual SAN related cmdlets fully supported (no longer a fling) for managing Virtual SAN
In Virtual SAN 5.5 in order to remove a disk/disk group without data lost, hosts were placed in maintenance mode with the full data evacuation mode from all disk/disk groups.
Virtual SAN 6.0 Introduces the support and ability to evacuate data from individual disk/disk groups before removing a disk/disk group from the Virtual SAN. So there is no need anymore to put a host in maintenance mode. Disks can be removed (hot swap) only with controllers with passthrough configuration
When a disk hits a permanent error, it can be challenging to find where that disk sits in the chassis to find and replace it. When SSD or MD encounters a permanent error, VSAN automatically turns the disk LED on.
User might need to locate a disk so VSAN supports manually turning a SSD or MD LED on/off.
VSAN 6.0 introduces the What-if API. This allows users to see what the result is of applying a certain storage policy to the available capacity.
Virtual SAN might need to resync data in certain conditions like a host failure. To be able to be able to keep the set compliancy for redundancy, a new replica needs to be built. In VSAN 5.5 there was no way to monitor the progress of the resynchronization of data. VSAN 6 has a dashboard which shows how much bytes are left to resync and also an expected time for the resync to be finished (ETA to compliance)
control over Rebalance operation
The rebalance operation in Virtual SAN is designed to proactively re-distribute data throughout the cluster in order to maintain a balanced consumption and distribution of storage capacity. By default, the rebalance operation is automatically activated whenever the storage capacity of the magnetic disks reaches 80% of utilization. The purpose of the operation is to distribute the data evenly throughout the cluster until the storage utilization is below the system’s defined threshold of 80%.
In VSAN 6 admins now can perform manual a rebalance by using the Ruby console by the command
vsan.proactive_rebalance –start ~/computers/cluster
More info in this very interesting whitepaper of VMware.
NexentaConnect for VMware Virtual SAN
The ability of VSAN is providing storage for virtual machines. It cannot share files for other purposes. However VSAN 6 supports NexentaConnect for VMware Virtual SAN which can be used to add NFSv3, NFSv4, and SMB file services to VMware Virtual SAN deployments by utilizing storage in host servers. This storage is managed through a vCenter console and can leverage security and domain integration, Active Directory Authentication and Authorization, and Kerberos support. NexentaConnect also provides inline compression and deduplication, IO path handling, and performance and health monitoring. The new solution is validated for virtual desktop workloads of up to 6,400 users and can provide folder and volume snapshots with replication and backup capabilities for disaster recovery.
A very comprehensive post about NexentaConnect for VMware Virtual SAN can be read here.
The What is new in VSAN 6.0 is here.