Storage on The IT Hollow

Atlantis USX with VMware VSAN?

Thu, 13 Mar 2014 12:45:12 +0000

Last week I attended the Virtualization Field Day 3 put on by the amazing staff at GestaltIT. One of the sessions was hosted by the folks at Atlantis Computing and they were giving us an overview of their Atlantis USX product.

All travel expenses and incidentals were paid for by Gestalt IT to attend Virtual Field Day 3. This was the only compensation given.

Atlantis USX Overview

Before we get to the crux of the issue, a quick refresher on what Atlantis USX does. The USX product will utilize existing SAN, NAS and DAS and combine it with a server ram to do caching. The value proposition is that Atlantis USX can carve out RAM to be used as either a whole datastore (SUPER FAST) or combine it with existing storage and have it act as a cache. Keeping the cache so close to the processor without having to go across a bus or HBA, which can add additional latency, can be an important addition to a performance strapped storage solution.

Should You Consider Pure Storage as your Next Array?

Thu, 06 Mar 2014 07:24:20 +0000

If you are coming up on a storage refresh cycle soon, Pure Storage is worth taking a look at as your new storage array. I was fortunate enough to see them present their solution at Virtualization Field Day 3 this year and got a good look at their storage.

All travel expenses and incidentals were paid for by Gestalt IT to attend Virtual Field Day 3. This was the only compensation given.

PernixData FVP 1.5 Beta

Tue, 28 Jan 2014 14:50:50 +0000

Good news for all of you eagerly awaiting the next iteration of the PernixData FVP software. Version 1.5 is now in Beta and you can request the download for your own testing from the following link http://info.pernixdata.com/Betaprogram.

Disclosure: At the time of this writing I am a PernixPro which entitles me to early access to software, licenses or other merchandise. The thoughts expressed in this post are my own and have not been vetted by PernixData.

Microsoft's Resilient File System (ReFS)

Mon, 13 Jan 2014 14:30:52 +0000

Microsoft has a new file system designed to increase data integrity, scalability and availability called the Resilient File System (ReFS). This file system has leveraged many of the NTFS file system goodies and expanded them to make it more scalable and prevent corruptions. ReFS was released with Server 2012 and at the moment is designed for use with file shares. It cannot be used as a boot volume at the present time, but this file system seems poised to replace NTFS down the road.

Microsoft Storage Spaces

Mon, 06 Jan 2014 14:15:25 +0000

Microsoft Storage Spaces feature used to handle data redundancy, scalability and performance. Storage Spaces takes a set of Just a Bunch of Disks (JBOD) and pairs them together to allow for; either failures of a disk, gaining the performance of multiple spindles, or gaining the space of multiple disks. Traditionally this has all been handled by creating a Redundant Array of Independent Disks (RAID) group. Some examples of RAID would be:

Striping (RAID 0)
Mirroring (RAID 1)
Parity (RAID 5 or 6)

Storage Spaces create a similar type of RAID Group but then throw a virtual disk on top of them so that multiple types of stripes can be used on the same disks. For example, three physical disks can be put into a storage space. From there, three separate types of VDISKs can be created, Mirrored, Spanned and Parity can then be placed on the same set of disks with no issue. The diagram below shows an example.

CASL with Nimble Storage

Mon, 25 Nov 2013 13:30:54 +0000

I was fortunate enough to get to spend an hour with Dmitriy Sandler from Nimble Storage to see what all the fuss was about with their product and more specifically their Cache Accelerated Sequential Layout (CASL) File System.

Hardware Overview

Let’s cover some of the basics before we dive into CASL. The storage array comes fully loaded with all the bells and whistles, out of the box. All the software features are included with this iSCSI array and include items such as:

Understanding the SSD write performance cliff

Mon, 28 Oct 2013 13:00:59 +0000

Solid State drives are much faster than their spinning disk predecessors, but can also have performance degradation due to how they interact with Operating Systems.

Flash consists of blocks of data and those blocks are full of smaller items called pages. A typical SSD might have block sizes of 512KB and 4KB pages.

There are 3 statuses that a healthy page could be in on a flash disk.

Written to: Data from the OS has been written to the page.
Unwritten to: The page is free and available to be written to by the Operation System.
Invalid: The page has data in it, but is available to be overwritten by the Operating System.

Are you thin or thick? Where at?

Tue, 26 Mar 2013 13:47:47 +0000

I’m often asked about how to provision virtual machine disks. This almost always comes down to, “Should I use thick or thin disks?” and then “Should I do thin provisioning on the array or on the hypervisor?”

So here we go: Thin vs Thick

Thin provisioning:

Thin provisioned disks don’t allocate all of the space during the provisioning of the storage. Instead, they allocate the space on demand. This is a great way to get more bang for you buck out of your storage. Let’s take a closer look with an example.

Virtual Simulators

Tue, 19 Feb 2013 14:02:16 +0000

If you’re an engineer and you’re trying to get more experience with a variety of different storage devices, you might find yourself in a bit of a pickle. Most customers settle one one or two storage vendors and that’s it. So if you work for one of these companies you can learn EMC or Netapp, etc. I highly doubt your company would be interested in purchases a few different types of storage devices so that you can learn them as they are quite expensive.

Jumbo Frames

Tue, 11 Dec 2012 16:00:54 +0000

Jumbo frames can be useful to optimize IP networks, especially in storage networking. This post should help to explain why using jumbo frames can be useful.

I’m not Jumbo, I’m just big boned!

First, let’s define what we mean by the term jumbo frame. As you can imagine it’s bigger than a normal frame.

A Jumbo frame simply means any frame with an MTU larger than 1500 bytes. What exactly does that mean? To really understand that we need to look at an Ethernet frame. The diagram below shows a hastily thrown together Ethernet frame and most of the frame we’re not concerned with for this topic. Parts of the frame are used for determining where the frame is headed, where it came from and to make sure it arrived intact. The section we’re looking at is the “Data” or “Payload” section of the frame.

HP 3PAR for midrange business

Mon, 03 Dec 2012 16:44:06 +0000

HP Enterprise class storage has just entered the mid range market. Today HP announced the HP 3PAR StoreServ 7000 class which includes two devices; the HP 3PAR 7200 and the HP 3PAR 7400. The 7200 starts at $25k for the 2U device and the 7400 (seen below) is less than $40K for a 4U device.

I’m very excited about this announcement because now HP has a storage device with the features that everybody wants and it’s now affordable for a smaller sized organization. HP has seemingly targeted one of it’s own devices with this announcement (the HP EVA) since it has been very popular with the mid-range business. They’ve even included some tools to migrate data from the EVA to the new 3PAR. I seriously doubt that the EVA will entirely go away, but the new big brother is going to steal some of their thunder.

Software iSCSI load balancing in ESXi 5

Mon, 03 Dec 2012 14:14:56 +0000

When you team NICs together in ESXi 5 you can pick from a variety of load balancing techniques to determine how traffic should flow over the adapters. You might think that setting up software iSCSI initiators in ESXi would be done in a similar manner. Add a VMkernel to a vSwitch, add a couple of adapters and set a teamingfailover policy. It turns out that this is not the case. You could setup a software iSCSI initiator this way, but it won’t provide you the teaming or failover you’ve intended.

VMware Path Masking

Tue, 30 Oct 2012 13:15:34 +0000

I’ve written posts in the past regarding LUN masking on a storage array, but it is possible to mask a path directly from your vSphere environment. I feel that if at all possible the masking should be handled at array level because the array is closest to the disk. Let’s face it, if vSphere shouldn’t see a LUN for one reason or another, then why is the array presenting it in the first place?

HP Storage Comparisons (Sept. 2012)

Mon, 10 Sep 2012 12:00:26 +0000

I have been recently thrown into the world of HP Storage, and have been trying to learn all of the storage techniques that are in the HP product line. I noticed that I couldn’t find anything that really did a compare and contrast of each of the products so I started to put one together. Anything I couldn’t understand, I asked a great guy named Calvin Zito (@hpstorageguy) to give me a hand with. He was more than gracious so follow him on Twitter.

VMDirectPath I/O Basic Setup

Tue, 10 Jul 2012 00:46:36 +0000

While I was studying for the VCAP-DCA I realized that many people might not have access to a lab that includes the capability to do VMDirectPath I/O. My own lab is using nested ESXi hosts inside of VMware Workstation so I don’t have access to DirectPath either, but I was able to borrow some equipment in order to test my skills.

If you don’t have access to this type of equipment but want to study for the VCAP5-DCA, the below setup should suffice for you to learn it, as the setup is not very difficult.

Path Selection Policy with ALUA

Thu, 08 Mar 2012 13:58:05 +0000

It’s important to understand how VMware ESXi servers handle connections to their associated storage arrays.

If we look specifically with fibre channel fabrics, we have several multipathing options to be considered. There are three path selection policy (PSP) plugins that VMware uses natively to determine the I/O channel that data will travel over to the storage device.

Fixed Path
Most Recently Used (MRU)
Round Robin (RR)

Let’s look at some examples of the three PSPs we’ve mentioned and how they behave. The definitions come from the vSphere 5 storage guide found below.