So I submitted a session for VMworld

Posted by admin | Posted on 18-05-2012 | Comment (1)

Category : Storage, Storage & Virtualization

A first for me:

Session Title: Simulating Storage IO workloads within virtual environments

Session Type: Breakout Session

Session Abstract:

With Tier 1 applications becoming commonplace within virtual environments, the need to simulate the IO profiles of those workloads prior to deployment has become a concern for the teams that manage virtual infrastructures. As we have seen with each subsequent release of vSphere, more emphasis has been placed on increasing the efficiencies of storage operations, and offloading those operations from the ESXi hosts and onto the storage arrays themselves. CPU and Memory resource performance has continued to increase as their costs have decreased, yet the one area where we cannot easily or quickly make significant cost effective changes is the shared storage platform. Where these memory and CPU resources were once look upon as the chief culprits in resource contention, storage IO performance is taking over as the prime area of concern for many IT shops.

This breakout session will be a discussion about the IO profiles of a number of varying production workloads, and how we can use tools such as the VMware IO Analyzer to set a baseline for IO profiles and how they will perform on current or proposed storage platforms within our virtual infrastructure.

Submitting to Present at VMworld: US

Session Outline:

Session Outline: A. Overview of the history of storage related changes within the ESX/ESXi hypervisor platform throughout the years. B. A look at some of the common IO profiles for Tier 1 application workloads that are targets for virtualization from the SMB to Enterprise space ( Oracle DB, Exchange, Lotus Notes, SAP, SQL, Video Delivery, OLTP, etc) C. Discussion about the various tools available to simulate IO workloads within virtual environments with an emphasis on the VMware IO Analyzer tool and how it can provide benchmarks for storage and IO performance. D. A roundup of results from performance testing done on various storage arrays. E. Looking towards the future with improvements in vSphere 5.1 and beyond.

Key Takeaway 1: Identifying how a given IO profile or workload will perform on existing virtual infrastructure is paramount to determining if you should virtualize that workload.

Key Takeaway 2 :Storage performance is the key to higher consolidation ratios within virtual environments, and the ability to simulate various IO workloads within your virtual environment will provide IT teams with the means to make a best possible choice on storage system purchases

Key Takeaway 3: End users can utilize free tools such as the VMware IO Analyzer to perform IO workload analysis and benchmarking for their existing environments.

Technical Level: Technical

Tracks: Infrastructure

Topics: Storage

 

 

Defending the indefensible

Posted by admin | Posted on 18-05-2012 | Comments (0)

Category : Facepalm, Waste

And the hits just keep coming:

West Virginia will save money on training technicians because they’ll only have to learn how to fix one type of router, not multiple routers, said Strickling, Obama’s point man on broadband issues.

“West Virginia believes they have found the most economical solution by buying a single product and getting a substantial discount,” Strickling said.

We are in the very best of hands :rolleyes: I’m pretty sure that there are no additional support steps for Cisco IOS that vary greatly between the different router models. Sure there may be additional cards to support that would not be common across different platforms, but the operating system is uniform.

And as for the “discount”, uh there is no way in hell the cost of buying the appropriate router needed for each location vs buying one size fits all was cheaper. This guys just talking out of his ass.

As of last week, more than 300 routers remained boxed up in storage. The stored routers came with a five-year service warranty, so the state has already lost two years of free maintenance on the devices.

At Wednesday’s hearing, Strickling was asked why West Virginia purchased more routers than needed. State officials, he said, later learned that many sites already had suitable routers and fiber connections.

Of course they did, and I’d hazard a guess they were working fine and didn’t need to be upgraded at all.

Yeah, lets drop a router that can support 50,000 users into that library that has 6 computers in it.

This is why we can’t have nice things, your tax dollars at waste.

Posted by admin | Posted on 16-05-2012 | Comment (1)

Category : and Abuse

Caught the Speaking in Tech podcast this morning while I was out for a run and was just floored by this story about the State of West Virginia buying $24 million worth of Cisco 3945 routers at $22k each for use in schools, libraries, and rural double wide trailers.

“The WVDE asked if the size of the routers could vary based on the needs of a school,” said Liza Cordeiro, spokeswoman for the Department of Education. “At that time, it is our understanding that, for consistency and future expansion, the plan was to buy all the same size.”

Gianato said putting the same size router in every school was about “equal opportunity.”

“We wanted to make sure a student in McDowell County had the same opportunities as a student in Kanawha County or anywhere else,” he said. “A student in a school of 200 students should have the same opportunity as a student in a school with 2,000 students.”

Seriously, WTF? “Equal Opportunity” its a damn router idiot.

The routers alone cost the state $7,800 each, but “add-ons” — additional equipment that came with the devices — boosted the price tag by $14,800.

“It’s like buying a car,” Gianato said. “You get a lot of options with the car.”

An online Cisco retailer was selling new 3945 series routers for $5,800 last week. The routers have a list price of $13,000 each.

I can’t even begin to comprehend this level of stupidity and waste, and honestly the only thing I can think of is that a massive level of fraud was perpetrated onto the people of West Virginia. Everyone involved in this boondoggle needs to be fired. When you start looking into the details (like Cisco being mentioned by name in the spec request, and that there was no RFP) the level of malfeasance continues to grow.

“I’m not an expert on the technical side,” he said, “but these have all kinds of capabilities and applications.”

Understatement of the Century.

And the kicker,  the Cisco rep who sold the gear retired after that sale.

 

Drinking from the firehose

Posted by admin | Posted on 09-05-2012 | Comments (0)

Category : Sales Engineering, VMWare

A favorite movie from my high school years was Weird Al Yankovic’s UHF.  Seriously, if you have never seen it, there is a treasure trove of one liners and perfectly timed bit comedic performances. For the vast majority of the world, it was their first exposure to Michael Richards aka Kramer from Seinfeld. Now there is a scene in the film where Richards as “Stanley Spadowski” was given his own children’s TV Show, and as a reward for “finding the marble in the oatmeal” a young contestant won a drink from the fire hose. Seriously, for the last 3 weeks I’ve felt like that kid.

Making the transition from end user administration and computing to the world of RAM’s, Channels, partner exchanges, demos, briefings, weekly team calls, etc. has been an eye opener to say the least. Not to mention, you are trying to learn an entire companies product line is as little time as possible because the sales group  (inside, channel, direct, etc) all are chomping at the bit to get you out in front of customers as soon as possible. It’s a lot to take in. To say it’s overwhelming is an understatement, but it’s also exhilarating. So far no two days have been the same, and I’m getting to meet really smart people and I’m also getting to see an entire different side of IT. Working and implementing technology on a daily basis can tend to insulate you from the challenges that are faced from the side that actually has to pitch, sell, and support those systems.

Interoperability between vendors, support statements, target markets, and a whole host of new terminology and practices and procedures for daily operations are introduced. I think one of the most challenging things I’ve faced is how I craft an answer to a customer. You really have to choose your words wisely and take into account a plethora of information and apply it in a manner that reflects well on the company you represent, as well as yourself and your product line. It’s something that as a designer and administrator you may not necessarily have to give as much thought too.

So if blogging has been light, it’s because I’ve been pouring through slide decks, white papers, and tech manuals. All of this also while I try to craft my first session submittal for VMworld this year. Talk about biting off more than one can chew J

 

It makes you want to cry

Posted by admin | Posted on 04-05-2012 | Comments (0)

Category : Facepalm, Storage

From the everlong “what did you learn today” thread at ARS, comes this tale of woe. Remember to vet those shipping companies. Also, a single tie down for something like that will never cut it. Apparently there goes $180k of Netapp gear.

One crazy month. A new job, and I’m now a vExpert for 2012

Posted by admin | Posted on 16-04-2012 | Comments (0)

Category : Emulex, vExpert, VMWare

Tags:

Sunday rocked to say the least. I woke up to find out that I was selected as a VMware vExpert for 2012. I am very thankful to have been nominated and to have been given recognition by the VMware community for my contribution to the community at large. I really hope that over the next year I can contribute even further and make an even stronger impact.

Congratulations to the 400 or so VMware community members who also were awarded vExpert recognition for this year. It was very awesome to see 150 first time vExperts join me in this year’s freshmen class.

On top of that, I’m starting a new gig on the 23rd. After being with Toshiba for the last 6 ½ years primarily as their chief storage and virtualization engineer, I am moving on to take a position with Emulex as a Systems Engineer on their sales team. I will be covering the South West, including Nevada, Utah, Arizona, Colorado, Hawaii, and Southern California.

For some time I have wanted to transition into a Sales Engineering role. Over the last few months I had interviewed with several companies including Veeam, and HP as well as few local and national VAR’s. I am really looking forward to getting into a customer facing and a more evangelical role where I believe my background in the administration and design space will allow me to really help the customer focus on their needs.  Emulex is well positioned to allow me to interface with a large variety of companies including the major players like IMB, HP, Dell, EMC, etc. It also will allow me to stay involved in my two tech passions, Storage and Virtualization, as well as allow me to become more involved on the Converged infrastructure front.

So look for a lot of cool stuff to come from this space in the near future.

Marketing masquerading as “Tech Journalism”, Or how not to write a stealth press release

Posted by admin | Posted on 11-04-2012 | Comment (1)

Category : Storage, Storage & Virtualization, Tech Marketing, VMWare

Let me preface this by saying that this may come across as snarky and malicious, but that’s not the full intent. I’m using this as a simple illustration of what bothers me about tech journalism. There is a trend I continue to see in technical marketing that bothers me to no end, it’s the press release that masquerades as a legitimate piece of tech reporting.

Case in point, this piece by Mike Vizard at IT Business Edge. There is so much wrong in this piece that I felt compelled to address it.

One of the first storage vendors to add specific support for the vStorage APIs for Array Integration (VAAI) in VMware vSphere is Tintri, a provider of storage appliances that combine Flash memory and SATA drives in a way that is specifically optimized for the VMware file system.

Interesting. A company founded in 2008, that doesn’t even come out of stealth until March 2011, somehow is one of the first storage vendors to add support for VAAI which was first referenced by EMC Japan in June 2010, and announced by VMware with the release of vSphere 4.1 in July 2010.  Whats that, almost a year before Tintri even ships its first box?

And what of being one of the “first” to offer VAAI support? Well given that Tintri still doesn’t (it’s coming this summer), I’d say that claim is a little off. In reality,EMC, 3Par, and Netapp were the first groups to really offer VAAI support on their arrays. IBM followed up pretty quick on their XIV and Storwise platforms. Compellent, HP, etc. etc. all soon followed.The list was initially pretty small, but if you look at the HCL today its growing with most of the Enterprise players having support for the first 3 primitives, and most supporting all 5 with the release of vSphere5.

Trust me, tech journalism isn’t the only group out there that takes a press release and regurgitates it to appear as a researched article with “facts”, the political press has done this for years. It’s sloppy and lazy, and frankly we deserve better.

For the record, I love what Tintri is doing and wish I could get one of their rigs to play with. This is by no means an attack on them, there are some really scary smart people working on that team. I spent almost half an hour with Dr. Harty at VMWorld going over the product and getting some real insight into what Tintri wants to do with their systems.  Still, it doesn’t do them any benefit to make claims that are simply not correct.

 

Hyper-V vs VMware a response to Nate at TechopsGuys

Posted by admin | Posted on 07-04-2012 | Comments (0)

Category : Hyper-V, Tech Marketing, Virtualization, VMWare

I read this post by Nate over at TechOpsGuys about Hyper-V vs VMware. It’s a good read and Nate brings several salient points to discussing the challenges facing VMware in the future as Hyper-V comes into feature parity with VMware.

This was my response to his post, I’m thinking of digging a lot deeper into this, but I wanted to go all stream of conciousness and get this out while it was still fresh.

I think that sometimes it’s easy to let the environment you manage color your perception of how we think others are deploying and managing technology as a whole. I know that when I write about certain aspects of technology that I will lean heavily on the assets I manage on a daily basis, its my familiarity with those systems that allow me to speak with authority about them. I sometimes take for granted that other IT shops do things differently, or may see my approach as not going to work for them. It’s very easy to generalize and jump to conclusions.

I actually had a boss that would tell me that “cost is no issue” for every project I worked on. I knew full well that cost was the over-riding issue and would architect 3 different scenarios for each project, the “cost is no issue”, the “what I think we would really need to meet the projects needs”, and the budget solution. Invariably, the budget solution would be the one that management opted for and I’d have to make due with it. I think it really depends on the type of corporate culture at a given company as to whether cost will be an issue or not. Some companies still love to hate their IT departments, and will do their best to make due with as little as possible and take the risk that they will survive a real disaster. Others will throw money at IT and waste it as they cluelessly look for a solution to a perceived problem. I know of a company that bought an XIV system to test a workload, and then pulled all the drives and shredded them after the test due to the sensitivity of the data and their internal policies. Yes, there is close to 800K down the drain, that’s the annual IT budget for lots of shops.

I can’t make a comparison of Hyper-V3 to VMware until it’s actually released. I refuse to take marketecture as fact, especially in the case of Microsoft products. I most certainly will not deploy any MS product until at least SP1 has come out.  Those burns refuse to heal.

We are 80/20 Windows/Linux so for us the benefits involved with memory management within VMware work well. Linux workloads, not so well, that may improve since VMware has Linux proficiency, will it with Hyper-V, I doubt it. We have achieved densities of around 35:1 with our existing host infrastructure. I don’t know of any Hyper-V shops currently getting that level with production class workloads.

I’ve yet to see a third party that I trust provide the cost comparison between Hyper-V and VMware. Those of us with MS ELA’s can get discounts at a significant level, but then there’s all the extra costs involved. Sure the hypervisor is free, but you want to manage it? Well that’s where its going to cost you plenty.

You’re right that VMware is expensive, and it’s going to get even more expensive as time goes on. The licensing fiasco for vSphere5 may have left people’s minds as time has gone by, but I tend to see it as VMware adopting some aspects of the Oracle pricing model. Oracle can charge what it does because they know they have many of their customers over a barrel. VMware is approaching that threshold with a fair number of large installs. Sure the SMB market may have the flexibility to move to Hyper-V, but your shops with 1000+ VM’s and all the infrastructure built out to support them will be hard pressed to make the move. This will mean that you will see shops keeping their older, and cheaper, versions of VMware unless there is a true benefit to upgrade.

If you look at the features being crafted with each subsequent release of vSphere, you will see that they are nearly all confined to IO performance, and Storage management/offload. In robust environments, this is where the pain point continues to manifest. In the general VMware populace at large, its memory constraints, but for those groups pushing the envelope, and running the 85-100% virtualized platform, IO and Storage Constraints are key. I think that’s why you see VMware doing what they are with the VAAI primitives, and Storage DRS.

As for Hyper-V, meh, good enough doesn’t cut it in my current organization. Then there is the lack of a real ecosystem devoted to Hyper-V and products that I can leverage. Sure as the market share increases, the ecosystem will develop, but really if I’m going to stake my job on production workloads and the SLA’s required, I’m not going to settle.

Bottom line, I don’t doubt that feature parity is coming, but is it too late in the game for that vast swath of VMware customer base to make the change mid-stream? Sure shops may go evaluate Hyper-V and use it for test/dev and in the case of some shops production. Still, given my own scenario, I don’t see Hyper-V making any headway in our organization, and as such, for many shops it will be a decision they ultimately have to make.

I do think that at some point the hypervisor will be given away free, it essentially is now, it’s the management that will cost you. So if you want to run 100 ESXi hosts of vSpher5 with 64GB of RAM and two sockets, then get crafty with scripting, you can essentially run a hypervisor cost of zero. Still the amount of time to manage and craft that solution may cost you as much, if not more than paying for the licenses.

 

VSPEX – The rebirth of the Mainframe/Mini in x86 form?

Posted by admin | Posted on 01-04-2012 | Comment (1)

Category : Storage & Virtualization, Tech Marketing, Virtualization, VMWare

EMC has trademarked a new product name: VSPeX. So what the heck is it?

Alternative: slow news day at El Reg.

It doesn’t take a genius to realize that with EMC’s hardware and coding chops, and direct line into VMware that purpose built virtual appliance coupled with the appropriate connectivity and storage are the next evolution of their business model. There is a lot more competition in the storage space these days, sure EMC is the top dog, but a lot of that is built of their dominance since the Sym days. As the other players catch up and surpass EMC in terms of functionality and value add, they will need to differentiate themselves and in turn create an even larger revenue model.

If you think about it, a converged platform, other than VCE/VBlock etc makes sense since it can be bolted into their existing storage product line. Just look at what they are doing with Greenplum, its the precursor to the VM appliance. Imagine a VMAX with built in ESXi and you have essentially the rebirth of the mainframe market for the x86 space. Scale on demand type delivery, as well as built in replication/dr, snaps, clones, etc. etc. Converge all the features of Enterprise storage and virtualization into an appliance coupled with their huge install base, and you have what Oracle wants to be but is incapable of becoming (since they unceremoniously killed SUN).

Virtrualization has been around for a long time, many of the features we use within the x86 virtrualization space have been available in the Mainframe/Mini/Unix space for some time.

Now the big one, does EMC go after Cisco? The long rumor has always been that EMC wants to buy a network stack, is Cisco too expensive? Personally, I wouldn’t go that route, I’d look at Extreme Networks given their Diamond series is the shit, and it could be had for a far cheaper price point without all the baggage of merging a huge company like Cisco into an equally huge company like EMC.

Of course, I could be crazy in thinking this. Wouldn’t be the first time.

IO Analyzer 56k no way!

Posted by admin | Posted on 21-03-2012 | Comments (2)

Category : Storage, Storage & Virtualization, VMWare

Install and configure VMware IO Analyzer:

Quick note, in case you missed the Brown Bag last night, the Video is here. I go over the install process as well as some general QA about IO Analyzer there.

VMware IO Analyzer is a VMware labs fling that was created to provide a simple analysis tool for testing storage attached to your VMware environment. For anyone who has installed an OVF template before, the Installation and setup is relatively simple.

Once there submit to the terms and conditions and you can download the zipped template:

 

 

 

 

 

 

 

 

Once the file has been saved to your local or network drive, expand its contents and open your vClient. We are going to want to go to File, Deploy OVF Template:

 

 

 

 

 

 

 

Browse to the templates location, and then click Next

 

 

 

 

 

 

 

You can now choose a new name for the VM or leave it as listed, also choose a folder location to deploy the VM if you have folders created within your environment:

 

 

 

 

 

 

 

Now choose the Cluster or Host you wish to deploy the VM onto:

 

 

 

 

 

 

 

 

Next if you wish to place the VM into a specific Resource Pool you can choose to do so here. I would not recommend doing so at this time, since we do not want to limit the resources available to the VM . Furthermore the resources utilized by the IO Analyzer are relatively small.

 

 

 

 

 

 

 

Now choose the data store you wish to place the VM onto:

 

 

 

 

 

 

 

You will want to install the VM as a thick provisioned device.

 

 

 

 

 

 

 

Choose the network you wish to access the IO Analyzer from:

 

 

 

 

 

 

 

 

Last step is to accept the final results and perform the installation:

 

 

 

 

 

 

 

 

The new VM will be created and placed into the location you provided above:

 

 

 

 

 

Once finished, we will want to configure the new IO Analyzer system for use.  Right click on the newly created VM and lets go to Edit Settings:

 

 

 

 

 

As you can see, the OVA comes with a pre-configured second disk of 100MB in size. If you consult the documentation within the IO Analyzer you will notice that the recommendation is to remove that disk and provide a new disk for testing.

 

 

 

 

 

 

 

 

 

I choose to utilize RDM (Raw Data Mappings) for this purpose. It can be done with either iSCSI or FC storage. In my case we have FC and will create a second disk for use on the system. This will be the test disk.  Prior to completing the next few steps you will need to follow your storage manufactures best practices to create a disk lun to present to the IO Analyzer virtual machine. For testing purposes, the lun should be over 2GB in size. This way we won’t utilize any of the cache on the IO Analyzer machine itself.

 

 

 

 

 

 

 

Add the new disk:

 

 

 

 

 

 

Choose Raw Device Mappings:

 

 

 

 

 

 

 

Remember you will need to re-scan your hosts storage prior to attaching any RDM devices.

 

 

Once you are sure the host can see your RDM disk then go ahead and add the disk to the IO Analyzer system.

 

 

 

 

 

Depending on your internal practices, you can store the disk with the VM or specify a datastore of your choosing. I tend to choose to store the RDM with the VM that its attached to.

 

 

 

 

 

 

 

 

You have the choice to either set the disk into Physical or Virtual lun mappings. This makes no difference as far as the test results are concerned.  If for any reason you wanted to snapshot the IO Analyzer, a virtual mapping would be required.

 

 

 

 

 

Following my own internal best practices, I will choose to attach the disk to its own separate SCSI Node:

 

 

 

 

 

Click finish and your new disk will be available for use:

 

 

 

 

 

 

 

 

 

One key thing to note, prior to launching the IO Analyzer is that you want to eagerzerothick the disk prior to actually running any tests on it. Follow the steps in this KB Article:

 

 

 

Now we are ready to power on the IO Analyzer and configure it for first use. Power on the machine, and lets go to the console to setup the networking  components. When the IO Analyzer first powers up, it will be sitting at this screen. Use the arrow buttons to move to the Configure Network section

 

 

 

 

 

 

If you use DHCP you can follow the prompts and your IP address will be configured. I would recommend against using DHCP for this system as you will need to access the web console via its IP, unless you plan on adding the system into your internal DNS structure.

 

 

 

 

Follow the prompts and enter the IP, Subnet, DNS and hostname information. Once you are satisfied, save your settings and the system will present with the new IP address showing at the top of the screen:

 

 

 

 

 

 

The next step is to logon to the system from the console. In order to perform any tests with the IO Analyzer, the system needs to be logged into. The credentials are root/vmware. After the logon, you will be presented with a rather bland linux gui:

 

 

 

 

 

 

 

 

At this point you are done configuring the system.  We can come back to this screen for more advanced settings in the future.

Now we can start running our first IO Analyzer tests. Open a firefox browser and lets input the IP address of the IO Analyzer machine.

 

 

 

 

As you can see there is a web based front end running that will provide the options to setup and run the IO Analyzer tests as well as view results. In order to get the proper ESXTOP results for our IO Analyzer as well as to select the IO Analyzer from the VM drop down menu, we want to add the ESX/ESXi host ip and the root password that the IO Analyzer is located on.

 

 

 

 

 

Next we will want to add the IO Analyzer system as a guest worker and create a workload to run our initial test on.

 

 

 

On this screen we several options. In this case I have a host, with an IO Analyzer system to choose from.  The fields are pretty intuitive, with the last one being the most important, that will contain the IP address of the IO Analyzer machine we want to test.  The one nice thing about the system is that once you setup your configuration, you can save it, and easily come back to run more tests in the future. See the Load and Save configuration tabs above.

So essentially, we are choosing the IO Analyzer Host, the IO Analyzer Virtual Machine, the test workload (which there are roughly 30 of) and we are inputting the IP Address of the IO Analyzer machine. Lastly we will add the worker and input a run time for the test. The last step is to click run.

 

Now it may appear that nothing is happening, but if we switch over the console of the IO Analyzer VM we can see the test as its running:

 

 

 

 

 

 

 

The IOmeter that is running on the IO Analyzer is fully interactive. You will have to slide the Update Frequency (red arrow above) over to a quicker refresh setting to see real time results. You can also alter or change the different tabs being displayed. This will be available during the entire run of the test.

When the test is over, the IO Analyzer web page will have the results available.

 

 

And the results link:

 

 

 

If configured properly you will see the IOmeter Results Summary, as well as the ESXTOP stats for the host being accessed.

And that’s it. For myself, I like to do independent verification of results, so I will look at my Veeam One  (shameful plug to my friends at Veeam, hey it’s free, you might as well try it) settings during and after the tests, as well as the monitoring tools on my storage array. You can also use the performance tabs on your ESXi hosts. I have verified that the results being provided are in line with what IO Analyzer is reporting.

 

 

 

 

 

 

 

 

 

 

 

Hope this helps, don’t forget to checkout the Brown Bag session I’m doing on this tonight 5 PM PST over at Professional VMware. If you have any questions comments hit me up.

Previous

  • Archives

  • The Twitters