VMworld 2013

Hi Friends,

I'm back from VMworld 2013 and I have to say, wow!  It's the busiest I've ever seen it.  We got a TON of foot traffic at the Nimble booth.  If you're still there today be sure to check it out!

We've got a ton of videos and demos and a SmartStack for you to look at and play around with if you want!

Here's a picture of me and my sister Rachel.  Notice the family resemblance?  We're fraternal twins. :-)  If you haven't been to Rachel's blog lately, you're missing out on some cool stuff!  Take a look!

As we were leaving the Moscone Center we saw the coolness of coolnesses.  A BACON truck!

That's all for now.  And if you missed it the first time around, here's my poem,

The Night Before VMworld 

Until Next Time

-Brain

VMware Horizon View - To SESparse or Not to SESparse, Here is The Paper!

Hi Friends,

In my recent blog, VMware Horizon View - To SESparse or Not to SESparse, That is The Question!
I told you a paper explaining our results and testing methodology was on it's way.  You can get it here:

https://connect.nimblestorage.com/docs/DOC-1291


I hope you enjoy reading it!

Until Next Time
-Brain

Cisco UCS 101 - VLAN's made easy

Hi Friends,

If you're like me you know just enough networking to be dangerous.  Sure I can log into a Cisco switch and add new machines to an existing VLAN and other simple things, but create a VLAN?  I'd have to do some Googling to do that!  I've been working with a Cisco UCS and I have to say this thing is VERY cool!  When you're working in a lab environment you'll soon discover you start bumping into your neighbors, accidentally use their DHCP addresses, AD server etc.  You can create new private IP ranges, but since you're on the same VLAN you'll start effecting each others work.

An easy way to segment your work off from your neighbors is to create a VLAN that your servers and storage can be dedicated to.  In a traditional world that involves tracing cables to ports, creating new VLANs on the switch, you know time consuming stuff!  Today I'm going to show you how easy it is to create a VLAN on a Cisco UCS.

1.  Log into your UCS Manager and click on the LAN tab.

2.  Find the Appliances cloud and expand it.  Now remember your UCS will probably look different then mine, so keep that in mind.  Find VLANs, right click on it and select Create VLANs.

3.  Give your new VLAN a name and a number.  You can give a range too, but I'm just doing simple stuff so I just give mine a single number.  Click OK when you're happy.

4.  Guess what?  You just created a VLAN!  Now for all the GUI haters out there I know you could have done this in probably half the time, but I still think it's cool!

So what to do with this new VLAN?  If you already have some virtual NICs in your blades or rack mounts you can apply this new VLAN to one of the vNICs.  If you don't have any free vNICS that can be segmented off, you can create a new vNIC!  How cool is that?!  If I seem excited it's because I am.  When I was a Sys Admin if you wanted to do this it meant actually adding a card to the box!

Let's add this VLAN to an existing vNIC.

5.  Click on the Servers tab at the top of the UCS Manager.

6.  Expand the tree until you find the server and vNIC you want to add the VLAN to.

7.  On the right hand side of the UCS Manager find Modify VLANs and click it.

8.  A window will open with all of your VLANs.  Deselect the VLAN you no longer want to be part of and select the one you do.  Here I deselect hyperv and select test.  Click OK to finish.

There you go, that's it!  That server now has a VLAN separated from everything else unless it's on that same 100 VLAN.

Until Next Time!
-Brain

DHCP Madness! - 0x800f0922 Error

Hi Friends,

I was configuring my DHCP server on Microsoft Windows 2012 and for some reason it stopped working.  When I tried to start it I would get a error that it couldn't be started.  So I thought I'd uninstall the role and reinstall, right?  The uninstall went fine, but when I tried to re-install I would get the following error:

Installation of one or more roles role services or features failed 0x800f0922

Hmm, having NO idea what this error meant I went to my friend Google and started searching.

I found a few instances of some others having this problem, and their solution was to disconnect the network because Windows seems to be trying to get something from the Internet.  I tried this, but still the same error.

I looked around more and found this discussion:

Unable to install DHCP server role on Windows 2012 Server Essential

It had similar answers, but I found one that said to go to C:\windows\system32\dhcp and delete the files within, but leave the directory.  I tried this and it worked!

So if you're having trouble with your DHCP server check out the link and maybe this solution will help!

-Brain

The Brain & Wendel Show #2 - SSD's, Flash and Cache

Hi Friends,

I'm happy to announce the second Podcast from the Brain and Wendel Show! 
 http://ow.ly/o08DH

For those without Apple products, go here to download the files directly.
http://www.mrkieran.com/Podcasts/podcast.xml


Wen and I talk about SSD's, Flash and Cache.

We hope you enjoy it!

-Brain

VMware Horizon View - To SESparse or Not to SESparse, That is The Question!

Hi Friends,

If you've been reading my blog for awhile now you'll know I've talked about SESparse vs. VMFSsparse and if we can believe what's written on the box we could certainly have a winner!  A little overview for those who haven't been long time Glick's Gray Matter readers.  In VMware Horizion View 5.1, VMware introduced a new format called Space Efficient Sparse Virtual Disks.  The problem in the past was that View wrote to disk in 512 byte chunks.  Now if the virtual desktops wrote everything in 512 byte chunks, there would be no problems, but they don't.  Since writes are variable and can be 512b, 4k, 16k, 32k, 64k, etc. we can start to misalign our storage.  Take a look at this diagram.

As you can see I've drawn out how VMFSsparse writes to disks.  Now depending on what boundaries your storage uses this diagram will vary.  Here I'm using a 4K block boundary and all would be cool if only 512b chucks of data were written to it.  In the drawing I've got 1 512b chunk of data written and then 4K.  Since the storage example above has 4K boundaries a piece of our 4K chunk has to be written to another block across another 4K boundary.  Now the problem occurs when I go to read that data.  Since it written across two blocks I have to do double the work on the storage.  This isn't too big of a deal when you just have a few desktops, but multiply this problem by 1000's and you can see this begins to really put extra demands on the storage.

In the past I thought there was an easy way to use or not use SESparse vs. VMFSsparse, but it looks like in 5.2 it's always on if certain environment variables are met.  If anyone from VMware is out there and I'm wrong, please let me know!  According to this article, SESparse can not be turned off from the GUI and must be done on the Horizon View server itself.  http://myvirtualcloud.net/?p=4745 

We ran some tests with SEsparse, with VMFSsparse and with View Storage Accelerator.  Here are our results:

The top line is average IO the desktops ran at steady state, ~8 IOPS, the next two blue lines are two tests running VMFSsparse and the two red lines are SEsparse.  You can see the number of misaligned I/O is about double for VMFSsparse!

If you'd like to learn more, we have a paper coming out shortly that will discuss this in further detail or you can talk to us about it at VMworld, just drop by the Nimble Storage booth and we'd be happy to chat to you about it!

Until Next Time
-Brain

Adventures in Technology - Spam!

Okay, for the legal mumbo-jumbo.  I'm am neither advocating nor condoning violence.  I'm a nice guy and would never hit people on the head with heavy objects and neither should you.  This is just a comic and it shouldn't be a guide on how to act towards people.  But it's kinda funny to watch stick figures abusing each other cause we shouldn't do it in real life!  So with that said, these comics should not be viewed by anyone..

How to Upgrade Your Nimble - It Can't Be That Easy! Can It?

Hi Friends,

Today I'm going to walk you through a process System Administrators avoid like the plague!  No, I'm not talking about a hair cut or a shave, I'm talking about upgrades.  Mention upgrades, patches, etc. and the hair on the back of their necks stand up!  Why?  Because upgrading sucks!  You have to arrange for downtime and NO one wants downtime even during maintenance windows.  The reason?  The power of centralized storage is everything is centralized.  The problem with it is everything is centralized...  When you bring it down, EVERYTHING on it stops.  Disruptive upgrades are a nightmare and that's when things go well!  I won't even get into what happens when things don't go well.

So with that said, I'd like to introduce you to the Nimble Storage upgrade.  Downtime?  NOPE!  Headaches?  NOPE!  Smooth sailing my friends.  There isn't downtime during an upgrade, what there is is a failover from the primary controller to the secondary controller with an ~30 second time.

How can this be?  Watch closely my friends and take in the goodness!

There are a couple of ways to do the upgrade, you can either download the newer Nimble OS from the Nimble Support Site directly from the administration GUI or if you don't have external Internet access you can download the Nimble OS from InfoSight and then upload it into the administration GUI.

Before you do anything, download and take a look at the release notes.  They can be found on InfoSight in the download area.  Pay special attention to the Verified Update Paths.  The last thing you want to do is make a jump that isn't supported.

I'm going to show you how to do an upgrade through an upload, so first I go to InfoSight and click on the Downloads link.

Select the Nimble OS tab, the version you want, here I'm choosing version 1.4.6.0, and click on Software.

Once that file downloads open up your Nimble GUI.  Click on Administration and then Software.  That will take you to the upgrade page.

You can see I'm currently on 1.4.3.  Click on the Upload button and find the update file.

Once you've located the file click on the Upload button.

The upload will begin and you'll be told when it's complete.  Remember this is NOT the upgrade, this is just uploading the new Nimble OS.  So if you don't feel comfortable doing this during business hours you can just upload the Nimble OS and update later.

You can see  here the downloaded Nimble OS is 1.4.6 and it's ready to Update when I am.  To update click on the Update button.

You can see here the update first installs on the secondary controller so if there are any issues, it won't be on your production controller.

Once Controller B has completed the upgrade the failover occurs and the upgrade begins on Controller A.

Once the Nimble OS has been loaded onto both controllers you're done!  You can see below the current version is now 1.4.6.  That's right, no pain, no reboot!

Sheesh, where was this when I was doing upgrades and patches?!  I would have more hair and what I have wouldn't be so gray!

Until Next Time

-Brain

Adventures in Technology - Backup

Okay, for the legal mumbo-jumbo.  I'm am neither advocating nor condoning violence.  I'm a nice guy and would never hit people on the head with heavy objects and neither should you.  This is just a comic and it shouldn't be a guide on how to act towards people.  But it's kinda funny to watch stick figures abusing each other cause we shouldn't do it in real life!  So with that said, these comics should not be viewed by anyone..

Adventures in Technology - I Won, I Won!

You asked for it, okay you didn't, but here it is anyway, Adventures in Technology!

Okay, for the legal mumbo-jumbo.  I'm am neither advocating nor condoning violence.  I'm a nice guy and would never hit people on the head with heavy objects and neither should you.  This is just a comic and it shouldn't be a guide on how to act towards people.  But it's kinda funny to watch stick figures abusing each other cause we shouldn't do it in real life!  So with that said, these comics should not be viewed by anyone..

Adventures In Technology!

Ever wanted to start your own comic strip, but people tell you you can't draw?  Well, I can't draw and I say %7##%^$($&^(*#%$^%#W on them!  Starting today I give you:

Adventures In Technology

Okay, for the legal mumbo-jumbo.  I'm am neither advocating nor condoning violence.  I'm a nice guy and would never hit people on the head with heavy objects and neither should you.  This is just a comic and it shouldn't be a guide on how to act towards people.  But it's kinda funny to watch stick figures abusing each other cause we shouldn't do it in real life!  So with that said, these comics should not be viewed by anyone...

 

Nimble Storage InfoSight - Empowering The Storage Administrator!

Hi Friends,

When I was a System and Storage Administrator, one of the most difficult things was to figure out if my systems were healthy and were they going to be healthy in 6 months.  When you're on-call, the last thing you want is to get a call at 2:30 in the morning that your storage system has run out of disk space.  This is something that is usually easy to fix, but will ruin your night and could halt production! 

There are a lot of great programs out there that will keep an eye on your servers, but what about your storage?  Wouldn't it be nice to have a webpage you could log in from anywhere and check the health and capacity of your storage?  Something that would show you ALL your Nimble Storage arrays?  Wow, I'm starting to sound like those infomercial guys...  Act now and you'll get two bottles for the price of one!  But I digress...

To answer my infomercial alter ego, YES, this page exists and it's available to you through InfoSight.  I've mentioned InfoSight a few times now and showed you some of the features, but I promised to come back and show the features in more detail.  Today I'm going to give you a tour of the Wellness and Capacity tabs.

Let's first take a look at the Wellness tab.  If you've been a System Administrator, you'll notice the information in the Wellness tab looks a lot like information found in syslog.  The nice thing is you don't have to go running around looking for these events, they're right at your finger tips.  Let's take a look at the image below.  From left to right.  The Event Summary shows messages from Urgent, Important to For Review.  Each are color coded, you can click on each to focus on those messages or go down the page to see the actual messages.  You can select Daily Summary Emails to receive information in your email about what you see here.  If you want helpful hints, select the Show hints box and you'll get lots of helpful bubbles that describe what the different features mean on the page.  The Event Details tell you what the Event Summary is alerting on.

Now let's look at the Capacity tab.  Capacity is really cool because it shows how much you're using on the array and provide predictive analysis on when the array will probably be filled.  This is great for future planning!  We've all gotten the emails from our boss that ask us what we need to purchase in the next few quarters.  With the predictive analysis, this makes it MUCH easier to plan for current and future growth.

Take a look where I've highlighted on the chart.  You can select any point and InfoSight will provide you with the information at that time.

I hope this was helpful and if it will help prevent one of those 2:30 in the morning phone calls, than my job is complete!  :-)

Until Next Time
-Brain

Occam's Razor Meet Neil.... Come Back Datastore!

Hi Friends,

I've been working with Nimble snapshots and I restored a previous snapshot of a volume that is a datastore in my ESXi 5.1 environment.  When you restore a snaphot Nimble has you take the volume offline.  It's a good idea so you don't shock the filesystem you're working with, in this case VMFS.

So no problem, I'm expecting my datastore to go offline as well as my virtual machines that are built from it.  I bring the volume back online after the snapshot restore and my datastore just continues to show Inactive...  No amount of refreshing, rescanning and pleading will bring it back!  The weird thing is it's showing active in the detailed storage adapter view.  I have it mounted on another ESXi server and it's showing showing up just fine!  What the heck?!

So I check the /var/log/vmkernel.log and I find, "Device is permanently unavailable".  Oh CRAP what the heck did I do?!  I'm thinking I've got some weird iSCSI duplicate ID's, or something weird.

Then I remember my friend Occam's Razor.  What's that you ask?  Basically all things considered, the simplest explanation is usually the answer.  I remember I was doing some CLI work on this ESXi machine and I check to see if I'm sitting in a directory that is no longer mounted on that machine.  Sure enough that's where I was!  I change directory out of there and guess what?  My datastore came back.  :-)

So what does this teach us?  NEVER use the CLI!   NO, no, just kidding.  If a problem seems totally bizarre and won't resolve itself, usually it's the simplest explanation.

Until Next Time
-Brain