NetApp has released the Virtual Storage Console 2.1 (VSC); our VMware vCenter plug-in. Behind the scenes the VSC is actually a framework from a number of sub-plug-ins (VSC, Rapid Cloning Utility and SMVI) delivered in a single installation. We created this sub-plug-in architecture as a way to enable new plug-ins without having to create multiple NetApp tabs in vCenter.
The main feature in the VSC 2.1 is a rewritten infrastructure discovery process. As I don't write code I am challenged to share with you the details of the changes made; however, I can share with you that a 30 node vSphere cluster, booting from SAN, with 1,505 virtual machines stored on 69 datastores used to take 35.45 minutes to discover. With version 2.1 the baseline discovery process now completes in 1.94 minutes and subsequent updates complete in 16 seconds!
The VSC also includes a new version of MBRAlign that supports I/O offload for correcting misaligned partitions in VMs running on NFS datastores. This capability reduces load on the hypervisor and storage network in order to reduce the time required to align the partition in the VMDK.
The VSC is the recommended best practice for deploying VMware on NetApp. It's also a free product available on the NOW site. Note, the backup and store capability requires a SMVI license.
Here's an overview of the capabilities of the VSC and enablments available in vCenter:
Supported Configurations
VMware VI3 & vSphere 4
- Independently and mixed clusters
- Data Ontap 7.3.3 and later (7.3.5+ & 8.0.1+ recommended)
- MetroCluster & vFiler storage options
Provisioning Storage
FC, FCoE, iSCSI & NFS datastores
- Create, clone, resize, and destroy
- Supports VAAI Full Copy & Block Zero
Secures access to storage
- LUN masking
- NFS exports
Cross-domain operation enablement
Storage admin creates storage pools
- VI admins provision from pools
- Role based access
- Clone VMs
- Provision, manage, destroy datastores
Storage Efficiencies
Enable & Monitor data deduplication
Enables hardware-based FlexClones of VMs & datastores
Desktop Virtualization
Automated pool creation
- Provisioning of hardware-accelerated VM clones & datastores
- Hardware-accelerated VM refresh
- VMware View Manager 4.6
- Citrix XenDesktop 5
OpenAPI Support
Provides programmable APIs for tech-centric customers & partners
SOAP based
Infrastructure Monitoring
Discovers arrays, hosts, and VMs
- Automatic update of storage controller list
- Audits against best practices established by NetApp & VMware
- Non-disruptively corrects items out of compliance
- Includes MBRTools
- Identify and correct VMDK misalignment
- I/O-offload for VMs on NFS
- Array details, disk type, data protection, replication, etc.
- Datastore, LUN, FlexVol, Aggregate
- Layers between datastore and aggregate are automatically managed
Backup & Restore
Agent-less Backup
Policy driven backups for datastores and VMs
Supports restores of datastores, VMs, VMDKs, & files
They say imitation is the sincerest form of flattery. If that's so, then color me flattered. This new NetApp plug-in is exactly like the EMC VSI 4.0 plug-in for vSphere that was released last year (http://akutz.wordpress.com/2010/12/22/emc-vsi-4-0/). I'm flattered, I truly am :)
Posted by: Schley Andrew Kutz | June 07, 2011 at 01:52 PM
@Andrew -
Thanks for sharing. I admit, EMC has done one fine job in developing their plug-ins and that is no short order with the number of disparate array platforms they support VMware on.
it's funny how EMC and NetApp employees view things differently. EMC claims 70+ points of integration scattered among 5 or 6 or 7 array platforms. I wonder what the median average is in terms of integrations per array platform?
NetApp provides the most complete plug-in, delivering the same capabilities across all arrays with access via any protocol. This model enables solutions consistency, regardless of being deployed in the core datacenter or in a small ROBO office. We think hardware is simply a means to scale.
EMC & NetApp view the needs of customers differently.
I challenge your claims that NetApp is not the leader and offer the following content from 2009 as evidence...
http://blogs.netapp.com/virtualstorageguy/2009/08/emc-the-storage-most-integrated-with-vmware-the-conclusion.html
Thanks again,
Vaughn
Posted by: Vaughn Stewart | June 08, 2011 at 02:23 PM
Does this replace SMVI? Should I not install that package and only use VSC 2.1?
Posted by: James Bothe | June 16, 2011 at 12:01 PM
Hi Vaughn,
Are there any more details about the new MBR Tools available other than what is in the IAG?
In an ESXi environment I would usually just use a Linux machine to mount the NFS stores and run the tool against the VMDK's as a part of any P2V process.
However the latest tool won't even let me run the help command on my Ubuntu 10.04 machine.
Is it totally redesigned to only work with ESX now? Any information would be helpful.
Kind Regards,
Jason.
Posted by: Jason Benedicic | June 17, 2011 at 09:13 AM
@Vaughn - a few of us have noticed the mbrscan and (the newer) nfsstat -d output are not agreeing on the state of alignment - can you weigh in on if this just an interpretation of nfsstat -d output issue or if we actually have unaligned IO happening on what mbrscan says are aligned VMs?
http://communities.netapp.com/message/56904
thanks
Posted by: www.facebook.com/profile.php?id=658313066 | June 22, 2011 at 10:46 AM
Hi!
When mbralign will support ESXi, if so?
How can I automate alignment for VMs on ESXi 4.1, just as mbralign does?
Thanks,
Pedro.
Posted by: Pedro Rocha | July 07, 2011 at 07:27 AM
@Fletcher - I don't have intimate knowledge around the data you are seeing with nfsstat, but I'd suggest you have an application in your VM that is creating a number of writes that are less than 4KB in size.
If what I suggest is accurate there's no need to be concerned, as it is normal behavior for the application.
I know you're well aware of misalignment, but please allow me to restate for those who may not be... What we want to avoid is having misaligned I/O for hundreds or thousands of VMs on an array. The inefficiency in I/O transfers due to a large mass of misaligned VMs will stress the array and lead to the eventual need to upgrade hardware.
Let me know if the small writes is or is not the case. I'd be happy to engage others to continue the conversation if needed.
Vaughn
Posted by: Vaughn Stewart | July 09, 2011 at 11:45 AM
@Pedro - We are woking on an official release targeted for the fall. Those pesky NDAs prevent me from sharing the details here.
If you need to address a misaligned ESXI environment today you should first upgrade to data Ontap 8.0.1. This release includes logic to better better handle the impact of misaligned workloads. After the upgrade you could consider a number of steps to actually align the VMDKs...
Wait for the MBRAlign releasee that supports ESXi
Running MBRAlign from an ESX host
Using a 3rd party alignment tool
Install MBRAlign in a Linux VM (btw - this model screams - its fast)
Posted by: Vaughn Stewart | July 09, 2011 at 11:50 AM