Category Archives: Storage

Top Posts 2016

2016 is pretty much done and dusted and it’s been an good year for Virtualization is Life! There was a more modest 70% increase in site visits this year compared to 2015 and a 2600% increase in visits since I began blogging in 2012. In 2016 I managed to produce 124 posts (including this one) which was slightly up on the 110 I produced in 2015 and in doing so passed 300 total blogs since I started here. I was fairly consistent in getting out at least eight blogs per month with June being my most prolific month with sixteen blog posts published.

Looking back through the statistics generate via JetPack, I’ve listed the Top 10 Blog Posts from the last 12 months. This year the opinion pieces seemed to be of interest to my readers and there is still vCloud Director and NSX representation in the top ten with my Veeam articles doing well. Again it was interesting to see that two of the most generic (older posts) and certainly basic posts took out two of the top three spots. It shows that bloggers should not be afraid of blogging around simple topics as there is an audience that will appreciate the content and get value out of the post.

  1. NSX Edge vs vShield Edge: Part 1 – Feature and Performance Matrix
  2. Quick Post: E1000 vs VMXNET3
  3. vSphere 6.0 vCenter Server Appliance: Upgrading from 5.x
  4. ESXi Bugs – VMware Can’t Keep Letting This Happen!
  5. Nutanix Buying PernixData: My Critical Analysis
  6. New NSX License Tier Thoughts and Transformers
  7. CBT Bugs – VMware Can’t Keep Letting This Happen!
  8. Veeam 9 Released: Top New Features
  9. Veeam’s Next Big Thing – Veeam has Arrived!
  10. vCloud Director 8: New Features And A New UI Addition…

I was honoured to have this blog voted #44 in the TopvBlog2016 and even with all the controversy around the voting I still hold that as a significant outcome of which I am very proud and I’d like to thank the readers and supporters of this blog for voting for me! And thanks must also go to my site sponsors who are all listed on the right hand side of this page.

With me moving across to vendor land it’s going to be interesting to see if I can keep up the variety of posts as I “narrow” down my core focus…however I fully intend to keep on pushing this blog by keeping it strong to it’s roots of vCloud Director and core VMware technologies like NSX and vSAN. I have the Home lab and the drive to continue to produce content around the things I am passionate about…and that includes all things hosting and cloud now with a touch of availability 🙂

Stay tuned for an even bigger 2017!

#LongLivevCD

Quick Look – vSphere 6.5 Storage Space Reclamation

One of the cool newly enabled features of vSphere 6.5 is the come back of VMFS storage space reclamation. This feature was enabled in a manual way for VMFS5 datastores and was able to be triggered when you free storage space inside a datastore when deleting or migrating a VM…or consolidate a snapshot. At a Guest OS level, storage space is freed when you delete files on a thinly provisioned VMDK and then exists as dead or stranded space. ESXi 6.5 supports automatic space reclamation (SCSI unmap) that originates from a VMFS datastore or a Guest OS…the mechanism reclaims unused space from VM disks that are thin provisioned.

When storage space is deleted without this automated feature the delete operation leaves blocks of unused space on the datastore. VMFS uses the SCSI unmap command to indicate to the array that the storage blocks contain deleted data, so that the array can unallocate these blocks.

On VMFS6 datastores, ESXi supports automatic asynchronous reclamation of free space. VMFS6 generally supports automatic space reclamation requests that generate from the guest operating systems, and passes these requests to the array. Many guest operating systems can send the unmap command and do not require any additional configuration. The guest operating systems that do not support automatic unmaps might require user intervention.

I was interested in seeing if this worked as advertised, so I went about formatting a new VMFS6 datastore with the default options via the Web Client as shown below:

Heading over the hosts command line I checked the reclamation config using the new esxcli namespace:

Through the Web Client you can only set the Reclamation Priority to None or Low, however through the esxcli command you can set that value to medium or high as well as low or none, but as I’ve literally just found out, these esxcli only settings don’t actually do anything in this release.

For the low setting in terms of reclaim priority and how long before the process kicks off on the datastore, the expectation is that any blocks that are no longer used will be reclaimed within 12 hours. I was keeping track of a couple of VMs and the datastore sizes in general and saw that after a day or so there was a difference in the available storage. 

You can see that I clawed back about 22GB and 14GB on both datastores in the first 24 hours. So my initial testing with this new feature shows that it’s a valued and welcomed edition to the new vSphere 6.5 release. I know that for Service Providers that thin provision but charge based on allocated storage, they will benefit greatly from this feature as it automates a mechanism that was complex at best in previous releases.

There is also a great section around UNMAP in the vSphere 6.5 Core Storage White Paper that’s literally just been released as well and can be found here:

References:

http://pubs.vmware.com/vsphere-65/topic/com.vmware.ICbase/PDF/vsphere-esxi-vcenter-server-65-storage-guide.pdf

https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2057513

vSphere 6.5 Core Storage White Paper Now Available

HomeLab – SuperMicro 5028D-TNT4 Storage Driver Performance Issues and Fix

Ok, i’ll admit it…i’ve had serious lab withdrawals since having to give up the awesome Zettagrid Labs. Having a lab to tinker with goes hand in hand with being able to generate tech related content…point and case, my new homelab got delivered on Monday and I have been working to get things setup so that I can deploy my new NestedESXi lab environment.

By way of an quick intro (longer first impression post to follow) I purchased a SuperMicro SYS-5028D-TN4T that I based off this TinkerTry Bundle which has become a very popular system for vExpert homelabers. It’s got an Intel Xeon D-1541 CPU and I loaded it up with 128GB or RAM. The system comes with an embedded Lynx Point AHCI Controller that allows up to six SATA devices and is listed on the VMware Compatibility Guide for ESXi 6.5.

The issue that I came across was to do with storage performance and the native driver that comes bundled with ESXi 6.5. With the release of vSphere 6.5 yesterday, the timing was perfect to install ESXI 6.5 and start to build my management VMs. I first noticed some issues when uploading the Windows 2016 ISO to the datastore with the ISO taking about 30 minutes to upload. From there I created a new VM and installed Windows…this took about two hours to complete which I knew was not as I had expected…especially with the datastore being a decent class SSD.

I created a new VM and kicked off a new install, but this time I opened ESXTOP to see what was going on, and as you can see from the screen shots below, the Kernel and disk write latencies where off the charts topping 2000ms and 700-1000ms respectivly…In throuput terms I was getting about 10-20MB/s when I should have been getting 400-500MB/s. 

ESXTOP was showing the VM with even worse write latency.

I thought to myself if I had bought a lemon of a storage controller and checked the Queue Depth of the card. It’s listed with a QD of 31 which isn’t horrible for a homelab so my attention turned to the driver. Again referencing the VMware Compatability Guide the listed driver for the conrtoller the device driver is listed as ahci version 3.0.22vmw.

I searched for the installed device driver modules and found that the one listed above was present, however there was also a native VMware device drive as well.

I confirmed that the storage controller was using the native VMware driver and went about disabling it as per this VMwareKB (thanks to @fbuechsel who pointed me in the right direction in the vExpert Slack Homelab Channel) as shown below.

After the host rebooted I checked to see if the storage controller was using the device driver listed in the compatability guide. As you can see below not only was it using that driver, but it was now showing the six HBA ports as opposed to just the one seen in the first snippet above.

I once again created a new VM and installed Windows and this time the install completed in a little under five minutes! Quiet a difference! Upon running a crystal disk mark I was now getting the expected speeds from the SSDs and things are moving along quiet nicely.

Hopefully this post saves anyone else who might by this, or other SuperMicro SuperServers some time and not get caught out by poor storage performance caused by the native VMware driver packaged with ESXi 6.5.


References
:

http://www.supermicro.com/products/system/midtower/5028/SYS-5028D-TN4T.cfm

https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2044993

Free Guide: Building NetApp ONTAP 9 Lab

In computing, there is one thing you shouldn’t compromise on…and that thing is storage. This carries over to Lab or NestedESXi environments as poor lab performance can be just as frustrating as production performance issues. I’ve used a number of nested storage platform’s for my lab environments and I’m always on the lookout for alternative solutions.

When Neil Anderson asked my to write a short introductory post on his new how-to guide Build Your Own NetApp ONTAP 9 Lab I decided to flick through the guide to check it out and see if it could add any value to my future plans for a homelab. The e-book is professionally laid out and has excellent diagrams, notes and step by step…it’s extremely comprehensive.

NetApp Simulator 9 Free eBook – Build Your Own NetApp ONTAP 9 Lab!

While I’ve never been a NetApp guy there does seem to be a level of complexity in the NetApp VSA setup, but with the step by step in the e-book any ambiguity is removed. If you are looking for lab storage this is a great end to end example of how to install and configure one based on ONTOP 9 NetApp Simulator.

Give it a look over here.

The Anatomy of a vBlog Part 1: Building a Blogging Platform

Earlier this week my good friend Matt Crape sent out a Tweet lamenting the fact that he was having issues uploading media to WordPress…shortly after that tweet went out Matt wasn’t short of Twitter and Slack vCommunity advice (follow the Twitter conversation below) and there where a number of options presented to Matt on how best to host his blogging site Matt That IT Guy.

Over the years I have seen that same question of “which platform is best” pop up a fair bit and thought it a perfect opportunity to dissect the anatomy of Virtualization is Life!. The answer to the specific question as to which blogging platform is best doesn’t have a wrong or right answer and like most things in life the platform that you use to host your blog is dependent on your own requirements and resources. For me, I’ve always believed in eating my own dog food and I’ve always liked total end to end control of sites that I run. So while, what I’m about to talk about worked for me…you might like to look at alternative options but feel free to borrow on my example as I do feel it gives bloggers full flexibility and control.

Brief History:

Virtualization is Life! started out as Hosting is Life! back in April of 2012 and I choose WordPress at the time mainly due to it’s relatively simple installation and ease of use. The site was hosted on a Windows Hosting Platform that I had built at Anittel, utilizing WebsitePanel on IIS7.5, running FastCGI to serve the PHP content. Server backend was hosted on a VMware ESX Cluster out of the Anittel Sydney Zones. The cost of running this site was approximately $10 US per month.

Tip: At this stage the site was effectively on a shared hosting platform which is a great way to start off as the costs should be low and maintenance and uptime should be included in the hosters SLA.

Migration to Zettagrid:

When I started at Zettagrid, I had a whole new class of virtual infrastructure at my hands and decided to migrate the blog to one of Zettagrid’s Virtual DataCenter products where I provisioned a vCloud Director vDC and created a vApp with a fresh Ubuntu VM inside. The migration from a Windows based system to Linux went smoother than I thought and I only had a few issues with some character maps after restoring the folder structure and database.

The VM it’s self is configured with the following hardware specs:

  • 2 vCPU (5GHz)
  • 4GB vRAM
  • 20GB Storage

As you can see above the actual usage pulled from vCloud Director shows you how little resource a VM with a single WordPress instance uses. That storage number actually represents the expanded size of a thin provisioned disk…actual used on the file system is less than 3GB, and that is with four and a half years and about 290 posts worth of media and database content  I’ll go through site optimizations in Part 2, but in reality the amount of resources required to get you started is small…though you have to consider the occasional burst in traffic and work in a buffer as I have done with my VM above.

The cost of running this Virtual Datacenter in Zettagrid is approx $120 US per month.

TipEven though I am using a vCloud Director vDC, given the small resource requirements initially needed a VPS or instance based service might be a better bet. Azure/AWS/Google all offer instance based VM instances, but a better bet might be a more boutique provider like DigitalOcean.

Networking and Security:

From a networking point of view I use the vShield/NSX Edge that is part of vCloud Director as my Gateway device. This handles all my DHCP, NAT and Firewall rules and is able to handle the site traffic with ease. If you want to look at what capabilities the vShield/NSX Edges can do, check out my NSX Edge vs vShield Series. Both the basic vShield Edges and NSX Edges have decent Load Balancing features that can be used in high availability situations if required.

As shown below I configured the Gateway rules from the Zettagrid MyAccount Page but could have used the vCloud Director UI. For a WordPress site, the following services should be configured at a minimum.

  • Web (HTTP)
  • Secure Web (HTTPS)
  • FTP (Locked down to only accept connections from specific IPs)
  • SSH (Locked down to only accept connections from specific IPs)

OS and Web Platform Details:

As mentioned above I choose Ubuntu as my OS of choice to run Wordpress though any Linux flavour would have done the trick. Choosing Linux over Windows obviously means you save on the Microsoft SPLA costs associated with hosting a Windows based OS…the savings should be around $20-$50 US a month right there. A Linux distro is a personal choice so as long as you can install the following modules it doesn’t really matter which one you use.

  • SSH
  • PHP
  • MySQL
  • Apache
  • HTOP

The only thing I would suggest is that you use a long term support distro as you don’t want to be stuck on a build that can’t be upgraded or patched to protect against vulnerability and exploits. Essentially I am running a traditional LAMP stack, which is Linux, Apache, MySQL and PHP built on a minimal install of Ubuntu with only SSH enabled. The upkeep and management of the OS and LAMP stack is not much and I would estimate that I have spent about five to ten hours a year since deploying the original server dealing with updates and maintenance. Apache as a web server still performs well enough for a single blog site, though I know many that made the switch to NGINX and use the LEMP Stack.

The last package on this list is a personal favorite of mine…HTOP is an interactive process viewer for Unix systems that can be installed with a quick apt-get install htop command. As shown below it has a detailed interface and is much better than trying to work through standard top.

TipIf you don’t want to deal with installing the OS or installing and configuring the LAMP packages, you can download a number of ready made appliances that contain the LAMP stack. Turnkey Linux offers a number of appliances that can be deployed in OVA format and have a ready made LAMP appliance as well as a ready made WordPress appliance.

That covers off the hosting and platform components of this blog…In Part 2 I will go through my WordPress install in a little more detail and look at themes and plugins as well as talk about how best to optimize a blogging site with the help of free caching and geo-distribution platforms.

References and Guides:

http://www.ubuntu.com/download/server

http://howtoubuntu.org/how-to-install-lamp-on-ubuntu

https://www.digitalocean.com/community/tutorials/how-to-install-linux-nginx-mysql-php-lemp-stack-in-ubuntu-16-04

VMworld 2016: Top Session Picks

VMworld 2016 is just around the corner (10 days and counting) and the theme this year is be_Tomorrow …which looks to build on the Ready for Any and Brave IT messages from the last couple of VMworld events. It’s a continuation of VMware’s call to arms to get themselves and their partners and customers prepared for the shift in the IT of tomorrow. This will be my fourth VMworld and I am looking forward to spending time networking with industry peers, walking around the Solutions Exchange on the look out out for the next Rubrik or Platform9 and attending Technical Sessions.

http://www.vmworld.com/uscatalog.jspa

The Content Catalog went live a few weeks ago and the Session Builder has also been live allowing attendees to lock in sessions. There are a total of 817 sessions this year, up from the 752 sessions last year. I’ve listed the main tracks with the numbers fairly similar to last year.

Cloud Native Applications (17)
End-User Computing (97)
Hybrid Cloud (63)
Partner Exchange @ VMworld (74)
Software-Defined Data Center (504)
Technology Deep Dives & Futures (22)

VMware’s core technology focus around VSAN and NSX again has the lions share of sessions this time year, with EUC still a very popular subject. It’s pleasing to see a lot of vCloud Air Network related sessions in the list (for a detailed look at the vCAN Sessions read my previous post) and there is a solid amount of Cloud Native Application content. Below are my top picks for this year:

  • Virtual SAN – Day 2 Operations [STO7534]
  • Advanced Network Services with NSX [NET7907]
  • A Day in the Life of a VSAN I/O [STO7875]
  • vSphere 6.x Host Resource Deep Dive [INF8430]
  • The Architectural Future of Network Virtualization [NET8193R]
  • Conducting a Successful Virtual SAN 6.2 Proof of Concept [STO7535]
  • How to design and implement VMware’s vCloud in production [SDDC9612-SPO]
  • PowerNSX and PyNSXv: Using PowerShell and Python for Automation and Management of VMware NSX for vSphere [NET7514]
  • Evolving the vSphere API for the Modern Era [INF8255]
  • Multisite Networking and Security with Cross-vCenter NSX: Part 2 [NET7861R]

My focus seems to have shifted back towards more vCloud Director and Network/Hybrid Cloud automation of late and it’s reflected in the choices above. Along side that I am also very interested to see how VMware position vCloud Air after the shambles of the past 12 months and I always I look forward to hearing from respected industry technical leads Frank Denneman, Chris Wahl and Duncan Epping as they give their perspective on storage and software defined datacenters and automation. This year I’m also looking at what the SABU Tech Marketing Team are up to around VSAN and VSAN futures.

As has also become tradition, there are a bunch of bloggers who put out their Top picks for VMworld…check out the links below for more insight into what’s going to be hot in Las Vegas this VMworld. Hope to catch up with as many community folk as possible while over so if you are interested in a chat, hit me up!

My top 15 VMworld sessions for 2016

Top 5 Log Insight VMworld Sessions

be_TOMORROW at VMworld 2016 – Key Storage and Availability Activities

 

My Top Session picks for VMworld 2016

http://www.mindthevirt.com/top-vmworld-sessions-category-1247

PowerCLI Script to Calculate VSAN vCAN Points Per Month

There is no doubt that new pricing introduced to vCAN Service Providers announced just after VSAN 6.2 was released meant that Service Providers looking at VSAN for their IaaS or MSP offerings that had previously written it off due to price, could once again consider it as a viable and price competitive option. As of writing this blog post there is no way to meter the new reporting mechanism automatically through the existing vCloud Usage Meter with the current 3.5 beta also lacking the ability to report billing info.

I had previously come across a post from @virten that contained a PowerCLI script to calculate VSPP points based on the original allocated GB model. With VSAN 6.2 pricing was now based on a consumed GB model which was a significant win for those pushing for a more competitive pricing structure to be able to push a now mature VSAN as a platform of choice.

Before I post the code it’s worth noting that I am still not 100% happy with the interpretation of the reporting:

The VsanSpaceUsage(vim.cluster.VsanSpaceUsage) data object has the following two properties which vCAN partners can use to pull Virtual SAN usage information: a) totalCapacityB (total Virtual SAN capacity in bytes) and b) freeCapacityB (free Virtual SAN capacity in bytes). Subtracting b) from a) should yield the desired “Used Capacity” information for monthly reporting.

I read that to say that you report for any fault tolerance or data resiliency overheads…that is to say if you have a VM with a 100GB hard disk consuming 50GB on a VSAN datastore utilizing RAID1 and an FTT=1 you will pay for the 100GB that is actually consumed.

With that in mind I had to add in a multiplier into the original script I had hacked together to cater for the fault tolerance and raid level you may run. The rest is pretty self explanatory and I have built on @virtens original script by asking for which vCenter you want to log into, what VSAN licensing model you are using and then finally ask for the RAID and FTT levels you are running. The result is the total amount of consumed storage of all VM disks residing on the VSAN datastore (which is the only value hard coded) and then the amount of vCAN points you would be up for per month with and without the overhead tax.

The code is below, please share and improve and note that I provide it as is and should be used as such. Please let me know if I’ve made any glaring mistakes…

If someone can also let me know how to round numbers and capture an incorrect vCenter login gracefully and exit that would be excellent! – [EDIT] Thanks to Virten for jumping on that! Code updated!

References:

PowerCLI Script to Calculate VSAN VSPP Points

VSAN 6.2: Reminder About Important Fix

[UPDATE] This issue is resolved in VMware ESXi 6.0, Patch Release ESXi600-201608001. For more information, see VMware ESXi 6.0, Patch Release ESXi600-201608001 (2145663).

Last week VMware released an important KB based around an issue with VSAN 6.2 where some VMs residing on existing Hybrid VSAN datastores may exhibit reduced disk IO performance after an upgrade. In a nutshell the issue is caused by a new operation that’s linked to the new deduplication and compression features in VSAN 6.2. The issue affects only VSAN 6.2 Hybrid deployments and is obviously not applicable to All Flash VSAN Clusters.

If impacted you may see:

  • A significantly lower than expected read cache hit ratio is observed on VSAN caching tier.
  • A higher percentage of IOPS may be observed on capacity tier disks on Hybrid diskgroups when compared from previous 6.x systems.
  • Overall increased VM observed latency

This issue is caused by VSAN 6.2 performing low level scanning for unique blocks, which is related to deduplication, can still occur on VSAN hybrid disk groups. This causes performance deterioration on Hybrid Disk groups, as it has a significant read caching performance impact on the SSD cache tier of VSAN disk groups.

The Workaround:

To work around this issue, if you are using a Hybrid configuration, you can turn off the dedup scanner option on each VSAN host in the VSAN Hybrid cluster. The way to turn it off is to modify the advanced setting lsomComponentDedupScanType which is set to a default value of 2. For the workaround you set that to 0. The easiest way to archive this is through PowerCLI as shown below.

Note that each host needs to be rebooted for the settings to take affect so go through the normal process of ensuring hosts go into VSAN maintenance mode before reboot.

Also worth mentioning a PowerCLI script that Jase McCarty has put up on GitHub that Gets/Sets the Deduplication Scanner settings with the use of some checks via a PowerCLI script that accepts variables.

https://github.com/jasemccarty/DedupeScan

References:

https://kb.vmware.com/kb/2146267

VSAN Upgrading from 6.1 to 6.2 Hybrid to All Flash – Part 3

When VSAN 6.2 was released earlier this year it came with new and enhanced features and with the price of SSDs continuing to fall and an expanding HCL it seems like All Flash instances are becoming more the norm and for those that have already deployed VSAN in a Hybrid configuration the temptation to upgrade to All Flash is certainly there. Duncan Epping has previously blogged the overview of migrating from Hybrid to All Flash so I wanted to expand on that post and go through the process in a little more detail. This is the final part of a three part blog series with the process overview outlined below.

Use the links below to page jump.

In part one I covered upgrading existing hosts, expanding an existing VSAN cluster and upgrading the license and disk format. In part two I covered the actual Hybrid to All Flash migration steps and in this last part I will finish off by going through the process of creating a new VSAN Policy, migrate existing VMs to the new policy and  then enable deduplication and compression.

Before continuing it’s worth pointing out that after the Hybrid to All Flash migration you are going to be left with an unbalanced VSAN cluster as the full data evacuation off the last Hybrid host will leave that host without objects. Any new objects created will work to re-balance the cluster however if you want to initiate a proactive re-balance you can tit the re-balance button from the Health status window. For more on this process check out this post from Cormac Hogan.

Create new Policy and Migrate VMs:

To take advantage of the new erasure coding now in the VSAN 6.2 All Flash cluster we will need to create a new storage policy and apply that policy to any existing VMs. In my case all VMs where on the Default VSAN Policy with FTT=1. The example below shows the creation of a new Storage Policy that uses RAID5 erasure coding with FTT=1. If you remember from previous posts the reason for expanding the cluster to four hosts was to cater for this specific policy.

To create the new Storage Policy head to VM Storage Policies from the Home page of the Web Client and click on Create New VM Storage Policy. Give policy a name, click Next and construct Rule-Set 1 which is based on VSAN. Select the Failure tolerance method and choose RAID-5/6 (Erasure Coding) – Capacity.

In this case with FTT=1 chosen RAID5 will be used. Clicking on Next should show that the existing VSAN datastore is compatible with the policy. With that done we can migrate existing VMs off the Default VSAN Policy onto the newly created one.

To get an list of what VMs are going to be migrated have a look at the PowerCLI commands below to get the VMs on the VSAN Datastore and then get their Storage Policy. The last command below gets a list of existing policies.

To apply the new Erasure Coding Storage Policy its handy to get the full name of the policy.

To migrate the VMs to the new policy you can either do it one by one via the Web Client of do it on mass via the following PowerCLI script.

Once run the VMs will have the new policy applied and VSAN will work in the background to get those VM objects compliant. You can see the status of Virtual Disk Placement in the Virtual SAN tab of the Monitor Tab of the cluster.

Enable DeDupe and Compression:

Before I go into the details…for a brilliant overview and explanation of DeDupe and Compression with VSAN 6.2 head to this post from Cormac Hogan. To enable this feature we need to double check that the licensing is correct as detailed in the first post and also ensure that all previous steps relating to the Hybrid to All Flah migration has taken place. To turn on this feature head to the General window under the Virtual SAN Settings menu on the cluster Manage tab and click on the Edit button next to Virtual SAN is Turned ON.

Choose Enabled in the drop down and take note of the checkbox that talks about Allow Reduced Redundancy understanding what that means by reading the info box as shown above. Once you click on the process to enable DeDuplication and Compress will begin…this process will go through an reconfigure all Disk Groups similar to to the process to upgrade from between Hybrid and All Flash. Again this will take some time depending on the number of host, number of disk groups and type of disks in the cluster.

Below I have shown the before and after of the Capacity window under the Virtual SAN tab in the Monitor section of the Cluster view. You can see that before enabled, there is a message saying that DeDeuplication and Compression is disabled.

And after enabling DeDuplication and Compression you start to get some statistics relating to both of them in the window relating to savings and ratios. Even in my small lab environment I started to see some benefits.

With that complete we have finished this series and have gone through all the steps in order to get to an All Flash VSAN Cluster with the newest features enabled.

References:

VSAN 6.2 Part 1 – Deduplication and Compression

VSAN 6.2 Part 2 – RAID-5 and RAID-6 configurations

 

VSAN Upgrading From 6.1 To 6.2 Hybrid To All Flash – Part 2

When VSAN 6.2 was released earlier this year it came with new and enhanced features and with the price of SSDs continuing to fall and an expanding HCL it seems like All Flash instances are becoming more the norm and for those that have already deployed VSAN in a Hybrid configuration the temptation to upgrade to All Flash is certainly there. Duncan Epping has previously blogged the overview of migrating from Hybrid to All Flash so I wanted to expand on that post and go through the process in a little more detail. This is part two of what is now a three part blog series with the process overview outlined below.

Use the links below to page jump.

In part one I covered upgrading existing hosts, expanding an existing VSAN cluster and upgrading the license and disk format. In this part am going to go through the simple task of extending the cluster by adding new All Flash Disk Groups on the host I added in part one and then go through the actual Hybrid to All Flash migration steps.

The configuration of the VSAN Cluster after the upgrade will be:

  • Four Host Cluster
  • vCenter 6.0.0 Update 2
  • ESXi 6.0.0 Update 2
  • One Disk Groups Per Host
  • 1x 480GB SSD Cache and 2x 1000GB SSD Capacity
  • VSAN Erasure Coding Raid 5 FTT=1
  • DeDuplication and Compression On

As mentioned in part one I added a new host to the cluster in order to give me some breathing room while doing the Hybrid to All Flash upgrade as we need to perform rolling maintenance on each hosts in the cluster in order to get to the All Flash configuration. Each host will be entered into maintenance mode and all data evacuated. Before the process is started on the initial three hosts lets go ahead and create a new All Flash Disk Group on the new hosts.

To create the new Disk Group head to Disk Management under the Virtual SAN section of the Manage Tab whilst the Cluster and click on the Create New Disk Group Button. As you can see below I have the option of selecting any of the flash devices claimed as being ok for VSAN.

After the disk selection is made and the disk group created, you can see below that there is now a mixed mode scenario happening where the All Flash host is participating in the VSAN Cluster and contributing to the capacity.

Upgrade Disk Group from Hybrid to All Flash:

Ok, now that there is some extra headroom the process to migrate the existing Hybrid Hosts over to All Flash can begin. Essentially what the process involves is placing the hosts in maintenance mode with a full data migration, deleting any existing Hybrid disk groups, removing the spinning disk, replacing them with flash and then finally creating new All Flash disk groups.

If you are not already aware about maintenance mode with VSAN then it’s worth reading over this VMware Blog Post to ensure you understand that using the VI Client is a big no no. In this case I wanted to do a full data migration which moves all VSAN components onto remaining hosts active in the cluster.

You can track this process by looking at the Resyncing Components section of the Virtual SAN Monitor Tab to see which objects are being copied to other hosts.

As you can see the new host is actively participating in the Hybrid mixed mode cluster now and taking objects.

Once the copy evacuation has completed we can now delete the existing disk groups on the host by highlights the disk group and clicking on the Remove Disk Group button. A warning appears telling us that data will be deleted and also lets us know how much data is currently on the disks. The previous step has ensured that there should be no data on the disk group and it should be safe to (still) select Full data migration and remove the disk group.

Do this for all existing Hybrid disk groups and once all disk groups have been deleted from the host you are ready to remove the existing spinning disks and replace them with flash disks. The only thing to ensure before attempting to claim the new SSDs is that they don’t have any previous partitions on them…if so you can use the ESXi Embedded Host Client to remove any existing partitions.

Warning: Again it’s worth mentioning that any full data data migration is going to take a fair amount of time depending on the consumed storage of your disk groups and the types of disks being used.

Repeat this process on all remaining hosts in the cluster with Hybrid disk groups until you have a full All Flash cluster as shown above. From here we are now able to take advantage of erasure coding, DeDuplication and compression…I will finish that off in part three of this series.

 

« Older Entries