PART I
Have you spent time searching the VMware documentation, on-line forums, venues and books to decide how to make a local dedicated direct attached storage (DAS) type device (e.g. SATA or SAS) be Raw Device Mappings (RDM)' Part two of this post looks at how to make an RDM using an internal SATA HDD.
Or how about how to make a Hybrid Hard disk drive (HHDD) that is faster than a regular Hard Disk Drive (HDD) on reads, however more capacity and less cost than a Solid State Device (SSD) actually appear to VMware as a SSD'
Recently I had these and some other questions and spent some time looking around, thus this post highlights some great information I have found for addressing the above VMware challenges and some others.
The SSD solution is via a post I found on fellow VMware vExpert Duncan Epping’s yellow-brick site which if you are into VMware or server virtualization in general, and particular a fan of high-availability in general or virtual specific, add Duncan’s site to your reading list. Duncan also has some great books to add to your bookshelves including VMware vSphere 5.1 Clustering Deepdive (Volume 1) and VMware vSphere 5 Clustering Technical Deepdive that you can find at Amazon.com.
Duncan’s post shows how to fake into thinking that a HDD was a SSD for testing or other purposes. Since I have some Seagate Momentus XT HHDDs that combine the capacity of a traditional HDD (and cost) with the read performance closer to a SSD (without the cost or capacity penalty), I was interested in trying Duncan’s tip (here is a link to his tip). Essential Duncan’s tip shows how to use esxcli storage nmp satp and esxcli storage core commands to make a non-SSD look like a SSD.
______________________________________________________________________
The commands that were used from the VMware shell per Duncan’s tip:
esxcli storage nmp satp rule add –satp VMW_SATP_LOCAL –device mpx.vmhba0:C0:T1:L0 –option “enable_local enable_ssd”
esxcli storage core claiming reclaim -d mpx.vmhba0:C0:T1:L0
esxcli storage core device list –device=mpx.vmhba0:C0:T1:L0
______________________________________________________________________
After all, if the HHDD is actually doing some of the work to boost and thus fool the OS or hypervisor that it is faster than a HDD, why not tell the OS or hypervisor in this case VMware ESX that it is a SSD. So far have not seen nor do I expect to notice anything different in terms of performance as that already occurred going from a 7,200RPM (7.2K) HDD to the HHDD.
If you know how to decide what type of a HDD or SSD a device is by reading its sense code and model number information, you will recognize the circled device as a Seagate Momentus XT HHDD. This particular model is Seagate Momentus XT II 750GB with 8GB SLC nand flash SSD memory integrated inside the 2.5-inch drive device.
Normally the Seagate HHDDs appear to the host operating system or whatever it is attached to as a Momentus 7200 RPM SATA type disk drive. Since there are not special device drivers, controllers, adapters or anything else, essentially the Momentus XT type HHDD are plug and play.
After a bit of time they start learning and caching things to boost read performance (read more about boosting read performance including Windows boot testing here).
Screen shot showing Seagate Momentus XT appearing as a SSD
Note that the HHDD (a Seagate Momentus XT II) is a 750GB 2.5” SATA drive that boost read performance with the current firmware. Seagate has hinted that there could be a future firmware version to enable write caching or optimization however, I have waited for a year.
Disclosure: Seagate gave me an evaluation copy of my first HHDD a couple of years ago and I then went on to buy several more from Amazon.com. I have not had a chance to try any Western Digital (WD) HHDDs yet, however I do have some of their HDDs. Perhaps I will hear something from them sometime in the future.
For those who are SSD fans or that actually have them, yes, I know SSD’s are faster all around and that is why I have some including in my Lenovo X1. Thus for write intensive go with a full SSD today if you can afford them as I have with my Lenovo X1 which enables me to save large files faster (less time waiting).
However if you want the best of both worlds for lab or other system that is doing more reads vs. writes as well as need as much capacity as possible without breaking the budget, check out the HHDDs.
Thanks for the great tip and information Duncan, in part II of this post, read how to make an RDM using an internal SATA HDD.
PART II
In the first part of this post I showed how to use a tip from Dunacn Epping to fake VMware into thinking that a HHDD (Hybrid Hard Disk Drive) was a SSD.
Now lets look at using a tip from Dave Warburton to make an internal SATA HDD into an RDM for one of my Windows-based VMs.
My challenge was that I have a VM with a guest that I wanted to have a Raw Device Mapping (RDM) internal SATA HDD accessible to it, expect the device was an internal SATA device. Given that using the standard tools and reading some of the material available, it would have been easy to give up and quit since the SATA device was not attached to an FC or iSCSI SAN (such as my Iomega IX4 I bought from Amazon.com).
Image of internal SATA drive being added as a RDM with vClient
Thanks to Dave’s great post that I found, I was able to create a RDM of an internal SATA drive, present it to the existing VM running Windows 7 ultimate and it is now happy, as am I.
Pay close attention to make sure that you get the correct device name for the steps in Dave’s post (link is here).
For the device that I wanted to use, the device name was:
______________________________________________________________________
From the ESX command line I found the device I wanted to use which is:
t10.ATA_____ST1500LM0032D9YH148_____Z110S6M5
Then I used the following ESX shell command per Dave’s tip to create an RDM of an internal SATA HDD:
vmkfstools -z /vmfs/devices/disks/ t10.ATA_____ST1500LM0032D9YH148_____Z110S6M5
/vmfs/volumes/datastore1/rdm_ST1500L.vmdk
______________________________________________________________________
Then the next steps were to update an existing VM using vSphere client to use the newly created RDM.
Hint, Pay very close attention to your device naming, along with what you name the RDM and where you find it. Also, recommend trying or practicing on a spare or scratch device first, if something is messed up. I practiced on a HDD used for moving files around and after doing the steps in Dave’s post, added the RDM to an existing VM, started the VM and accessed the HDD to verify all was fine (it was). After shutting down the VM, I removed the RDM from it as well as from ESX, and then created the real RDM.
As per Dave’s tip, vSphere Client did not recognize the RDM per say, however telling it to look at existing virtual disks, select browse the data stores, and low and behold, the RDM I was looking for was there. The following shows an example of using vSphere to add the new RDM to one of my existing VMs.
In case you are wondering, why I want to make a non SAN HDD as a RDM vs. doing something else' Simple, the HDD in question is a 1.5TB HDD that has backups on that I want to use as is. The HDD is also bit locker protected and I want the flexibility to remove the device if I have to being accessible via a non-VM based Windows system.
Image of my VMware server with internal RDM and other items
Could I have had accomplished the same thing using a USB attached device accessible to the VM'
Yes, and in fact that is how I do periodic updates to removable media (HDD using Seagate Goflex drives) where I am not as concerned about performance.
While I back up off-site to Rackspace and AWS clouds, I also have a local disk based backup, along with creating periodic full Gold or master off-site copies. The off-site copies are made to removable Seagate Goflex SATA drives using a USB to SATA Goflex cable. I also have the Goflex eSATA to SATA cable that comes in handy to quickly attach a SATA device to anything with an eSATA port including my Lenovo X1.
As a precaution, I used a different HDD that contained data I was not concerned about if something went wrong to test to the process before doing it with the drive containing backup data. Also as a precaution, the data on the backup drive is also backed up to removable media and to my cloud provider.
Thanks again to both Dave and Duncan for their great tips; I hope that you find these and other material on their sites as useful as I do.
Meanwhile, time to get some other things done, as well as continue looking for and finding good work a rounds and tricks to use in my various projects, drop me a note if you see something interesting.
Additional Information
Which Enterprise HDDs to use for a Content Server Platform
Obviously the data center virtualization is of importance for multiple reasons, but Horizon View as well.
I would consider our deployment, at least from the college’s deployment, vanilla, meaning we do not leverage a lot of the technologies VMware offers. We do deploy HA + DRS clustering, but that is about the extent of it.
Our vITA environment does have its uniqueness, and we continually attempt to develop labs that can address most of the products/features available from VMware.
VMware-based solutions are designed for the consolidation of servers. Also, since we had to expand our market globally to support the expense of running our vITA program, we had to come up with a delivery method to teach these courses anywhere in the world.
By using Horizon View's virtual desktop technology as the portal for participants to gain access to our virtual lab environment along with use of live online meeting tools (currently we use Adobe Connect), we became early adopters of the course delivery method now known as VILT (Virtual Instructor Led Training).
Continue to develop products that address the SMB market.
I have used VMware products for ten years.
My initial use was to teach Operating Systems at Caldwell Community College & Technical Institute. Within a year after I began using it for curriculum courses, Google decided to build its largest datacenter in the world just out our backdoor. We were invited by Google to develop a program to train individuals how to become “Datacenter Technicians”. I became intimately involved with this due to my industry background and my use of open source products, including VMware. Due to the rapid turnover in courses, preparation of VMware’s Workstation product became too time consuming so I installed the VMware Server solution, which at that time was v3.5.
Primarily since we were early adopters there was little expertise available, other than directly from VMware, which is one of VMware’s strong points in that they provide a wealth of information through their documentation (too much) and their community forums.
Hardware compatibility issues, in particularly early on, needed to be identified prior to attempting deployments. This is not really an issue with VMware products. Their guides refer back to verify compatibility with the HCL. Now most vendors ensure their hardware does comply. There were also issues arising from integration of vSphere with SAN vendor hardware. Again, most of these issues occurred early on due to our learning curve.
For the college, not only being “vanilla”, we are also not a huge institution so scalability is not an issue.
For our vITA program, we had to find ways to get the most from our available hardware. We initially had old equipment from the college as they increased the use of virtualization. I actually embrace this approach since I have been in the technology field for four decades. I consider it a challenge to get the most from limited resources. If you have ample resources, time and money, you should be able to accomplish most anything technologically. The skill/talent, at least from my point of view, is being able to accomplish this without the abundance of time/money/resources.
From the college, we have not had many occurrences to contact VMware support directly. Some of this is had to do with the relationship we had between myself, being the vITA Director, and the colleges Network/System Administrator. I did the research and development, which is basically what I have done both in my industry career and in my academic career, thus the college benefitted from my lumps on implementation on the production side.
With the vITA program, I was pretty much on our own, but did have access to some VMware internal information.
We didn't use any previous solution for server virtualization. For desktop, the college still uses XenWorks, with minimal Horizon View deployment mainly due to manpower issues and comfort.
We were early adopters, so obviously there were complexities.
We did it in-house.
From my point of view, particularly in the IT industry, you need to be continually moving forward, otherwise you are moving backwards or out. But that is not to say there is no room for improvement in particular areas, for instance, in addressing products that help the small business arena. With discussions I have had with internal VMware employees, they have known this and have introduced products, like VSAN, to help address this arena.
Get buy-in from other areas within your organization, which is typically an easy sell. But do it up front and identify a relatively small test deployment and the internal level of expertise. Then fill voids with either internal training or by establishing partnerships.
This is a logical diagram of our vITA Lab environment:
VMware has announced version 6 (V6) of its software defined data center (SDDC) server virtualization hypervisor called vSphere aka ESXi. In addition to a new version of its software defined server hypervisor along with companion software defined management and convergence tools.
As a refresh for those whose world does not revolve around VMware, vSphere and software defined data centers (believe it or not there are some who exist ;), ESXi is the hypervisor that virtualizes underlying physical machines (PM’s) known as hosts.
The path to software defined data center convergence
Guest operating systems (or other hypervisors using nesting) run as virtual machines (VM’s) on top of the vSphere hypervisor host (e.g. ESXi software). Various VMware management tools (or third-party) are used for managing the virtualized data center from initial configuration, configuration, conversion from physical to virtual (P2V) or virtual to virtual (V2V) along with data protection, performance, capacity planning across servers, storage and networks.
VMware vSphere is flexible and can adapt to different sized environments from small office home office (SOHO) or small SMB, to large SMB, SME, enterprise or cloud service provider. There is a free version of ESXi along with paid versions that include support and added management tool features. Besides the ESXi vSphere hypervisor, other commonly deployed modules include the vCenter administration along with Infrastructure Controller services platform among others. In addition, there are optional solution bundles to add support for virtual networking, cloud (public and private), data protection (backup/restore, replication, HA, BC, DR), big data among other capabilities.
VMware has streamlined the installation, configuration and deployment of vSphere along with associated tools which for smaller environments makes things simply easier. For the larger environments, having to do less means being able to do more in the same amount of time which results in cost savings. In addition to easier to use, deploy and configure, VMware has extended the scaling capabilities of vSphere in terms of scaling-out (larger clusters), scaling-up (more and larger servers), as well as scaling-down (smaller environments and ease of use).
The quick synopsis of VMware vVOL’s overview:
How data storage access and managed via VMware today (read more here)
vVOL’s are not LUN’s like regular block (e.g. DAS or SAN) storage that use SAS, iSCSI, FC, FCoE, IBA/SRP, nor are they NAS volumes like NFS mount points. Likewise vVOL’s are not accessed using any of the various object storage access methods mentioned above (e.g. AWS S3, Rest, CDMI, etc) instead they are an application specific implementation. For some of you this approach of an applications specific or unique storage access method may be new, perhaps revolutionary, otoh, some of you might be having a Deja Vu moment right about now.
vVOL is not a LUN in the context of what you may know and like (or hate, even if you have never worked with them), likewise it is not a NAS volume like you know (or have heard of), neither are they objects in the context of what you might have seen or heard such as S3 among others.
Keep in mind that what makes up a VMware virtual machine are the VMK, VMDK and some other files (shown in the figure below), and if enough information is known about where those blocks of data are or can be found, they can be worked upon. Also keep in mind that at least near-term, block is the lowest common denominator that all file systems and object repositories get built-up.
How VMware data storage accessed and managed with vVOLs (read more here)
Here is the thing, while vVOL’s will be accessible via a block interface such as iSCSI, FC or FCoE or for that matter, over Ethernet based IP using NFS. Think of these storage interfaces and access mechanisms as the general transport for how vSphere ESXi will communicate with the storage system (e.g. their data path) under vCenter management.
What is happening inside the storage system that will be presented back to ESXi will be different than a normal SCSI LUN contents and only understood by VMware hypervisor. ESXi will still tell the storage system what it wants to do including moving blocks of data. The storage system however will have more insight and awareness into the context of what those blocks of data mean. This is how the storage systems will be able to more closely integrate snapshots, replication, cloning and other functions by having awareness into which data to move, as opposed to moving or working with an entire LUN where a VMDK may live.
Keep in mind that the storage system will still function as it normally would, just think of vVOL as another or new personality and access mechanism used for VMware to communicate and manage storage. Watch for vVOL storage provider support from the who’s who of existing and startup storage system providers including Cisco, Dell, EMC, Fujitsu, HDS, HP, IBM, NetApp, Nimble and many others. Read more about Storage I/O fundamentals here and vVOLs here and here.
Depending on your experiences, you might use revolutionary to describe some of the VMware vSphere V6 features and functionalities. Otoh, if you have some Deja vu moments looking pragmatically at what VMware is delivering with V6 of vSphere executing on their vision, evolutionary might be more applicable. I will leave it up to you do decide if you are having a Deja vu moment and what that might pertain to, or if this is all new and revolutionary, or something more along the lines of technolutionary.
VMware continues to execute delivering on the Virtual Data Center aka Software Defined Data Center paradigm by increasing functionality, as well as enhancing existing capabilities with performance along with resiliency improvements. These abilities enable the aggregation of compute, storage, networking, management and policies for enabling a global virtual data center while supporting existing along with new emerging applications.
If you were not part of the beta to gain early hands-on experience with VMware vSphere V6 and associated technologies, download a copy to check it out as part of making your upgrade or migration plans.
Check out the various VMware resources including communities links here
VMware vSphere Hypervisor getting started and general vSphere information (including download)
VMware vSphere data sheet, compatibility guide along with speeds and feeds (size and other limits)
VMware Blogs and VMware vExpert page
Various fellow VMware vExpert blogs including among many others vsphere-land, scott lowe, virtuallyghetto and yellow-bricks among many others found at the vpad here.
StorageIO Out and About Update – VMworld 2014 (with Video)
VMware vVOL’s and storage I/O fundamentals (Storage I/O overview and vVOL, details Part I and Part II)
How many IOPs can a HDD or SSD do in a VMware environment (Part I and Part II)
VMware VSAN overview and primer, DIY converged software defined storage on a budget.
Overall VMware vSphere V6 has a great set of features that support both ease of management for small environments as well as the scaling needs of larger organizations.
Another great tip - if you use Nimble Storage install the Nimble Connection Manager software on your hosts for the pathing management. It works very well with these devices for connectivity.