Category: Releases

September sprint includes:

Budgetary Quote UI:

User doesn’t have to download excel to see the pricing on quote. The new interactive UI will allow one to see list prices and apply discounts to see the sales/net price. You can iterate the entire process (size, generate budgetary quote and apply discounts) and download the quote when fully satisfied.

Enhancements to the Financial assumptions:

– Support levels have been sorted

– Monthly support terms (14months, 26 months..etc.) have been added

– Brief description has been been added next to each support level

– Ability to apply discounts to Budgetary Quote has been moved to the new Budgetary Quote UI.

Regular/Robo Models

There was confusion on how robo models like the 1175S should be sized.  For Regular models which are used in data centers,  PM originally wanted to restrict the 1175S to just backup usage.  That in ROBO they could be used as application nodes that could be 1, 2, or 3+ nodes.

That brought some confusion when SE wanted to size a small cluster with say qty 5 of the 1175S.  Sure could do it in ROBO, but SE would say it is not a ROBO customer project.

So I worked with PM and we streamlined the rules and now models like the 1175S (lot of vendors have similar ones) are treated as fully functional models as follows:

Regular Model Rules

• All models included
• All use cases are allowed – main cluster application, remote cluster application and remote snapshots
• 3+ nodes are recommended for any model. It is considered good data center practice to have a minimum of 3 nodes.

ROBO Model Rules

• All models but only some models like the 1175S can size for 1 or 2 node, while others require 3 min nodes
• All use cases – main cluster application, remote cluster application and remote snapshots
• All models can go to 3+ nodes depending on sizing requirements

So what should you do

• Just stay with Regular models for most of your sizing needs. It is default and models like 1175S can indeed run applications.  Any recommendation would be 3+ nodes
• Only go to ROBO if you need 1 or 2 node for 1175S

For more info go to the Sizer wiki –  ROBO-Regular

Rack Awareness:

Nutanix has always had node awareness and for long time block awareness both of which are in Sizer.  With this release, Sizer will also have rack awareness where Data Availability is maintained even in an event of an entire rack or top-of-the-rack switch failure.

Supported Use Cases

• Heterogeneous solution made up of several different homogeneous blocks

–    For example, could have several 3460 blocks and several 3360 blocks

–    Here the 3460 blocks are all identical, while all the 3360 blocks are identical.  Given that, the 3460 block is homogeneous and the 3360 block is homogeneous.

–    These homogeneous blocks are dispersed across sufficient racks to meet rack awareness

• We will support a cluster wide setting for rack awareness and assume all workloads in that cluster must adhere to rack awareness

Other things

• Share notice in email. Sharing has been around forever and put in your Shared Scenarios but now when a person shares a scenario an email goes to them.  Great to increase collaboration and we will do more
• We have all Sizer and “non-Sizer” parts for all HP and Cisco models. Non-Sizer parts are things sizer does not analyze like say storage controller, boot drives , etc.  Where before we asked you to click link to HFCL we got in the BOM.
• Lots of product updates
• Inspur is a new vendor
• 3060-G6 NVME is there
• We allowed “no modification” as option for VDI when selecting desktop or office version.  Sometimes people have need to size for so much MHz or Cores per user and don’t want the modification.  In future Collector will use that option too when we upload a collector output

 

 

In August we added

• Collector 1.0 is now General Available (GA).   Nutanix Collector will ultimately change how Nutanix sells its products where we first collect all the pertinent customer workload data to then design the optimal solution in Sizer. 

In its first release, it takes inventory snapshots of the application VMs. In next few months it will take in time series information to get best insight.

VDI will be the first workload use case. Ultimately we will cover several different workloads, hypervisors and operating systems

• Collector Input.   Just like uploading a RVtools output into Sizer you can upload a Collector output

• Sizer Wiki –  This facility 🙂   This is important for the following reasons
  • All Sizer users (NTNX users and partners) get access to the information
  • It is safe as need to be a sizer user
  • Allows us to communicate updates and explain how Sizer works
• Planned failover for Storage Calculator.  Now you can select the failover level you expect and get the storage capacity for a cluster

• Numerous product updates across various vendors
• Hitachi added as a vendor

 

hi everyone

A few presents from the Sizer team before SKO. This is our July sprint

– NEC Vendor addition

– Common panel to define license, support and discounts. Just above the workloads on the left. This is where you apply these things for quotes, BOM, financial analysis

– Discounts support in Budgetary quotes – Now can apply discounts in in budgetary quotes. So you can see net price !!!

– Multiple nodes support in Storage capacity calculator – Hey if one node is cool why not more!

– License and Support SKUs support in BOM – we are getting it to be more complete

– Extent Store View in a scenario – We are trying to give you every chart known to mankind .. well at least storage people. Extent store is reported in sizing details and now a nice chart

– Replacing License/Support configuration in quote modals with data from financial assumptions

Key things added

• Unbalanced DIMM skylake update. Skylake is different than Broadwell for unbalanced dimms. With broadwell it was just dimm count. With skylake it is dimm count at times, but at times it also depends on how dimms are laid out on the motherboard. Finally the memory bandwidth can effect things too. At any rate we made big update here and Sizer takes all that into account.
• Storage Calculator updates. From user feedback we streamlined some things. For example the extent store is adjusted by the RF factor you set. So you see the combined storage but also breakout for ssd and hdd. We did add newer drives like the 10TB hdd
• Budgetary quote – now has license and support
• We re-enabled license and support with SFDC quotes. This was hard work in that we had a swat team that came up with lot of test cases and just kept ironing out issues.
  • Why you should care as an SE. Now go ahead and create the quote within sizer and then either you or your rep can deal with discounts and approvals in SFDC. The benefit is the quote lines are created properly. Both a labor savings and less errors

 

Here is more info on unbalanced dimms

Good news right now no model has the expensive 50% hit

Skylake processors is more complex compared to Broadwell in terms of whether a system has unbalanced dimms

We now test for the following

• CPU – skylake
• Model – Balanced_Motherboard – true (described below)
• Memory bandwidth – go with slower figure for either memory or CPU. If 2133 Mhz then -10% memory adjustment. If 2400Mhz or 2666Mhz (most common with skylake models) we take a 0% adjustment

Like before, we find the DIMM count per socket. There is typically 2 sockets (cpu’s) but can be 1 and starting to introduce 4 socket models

Using the quantity of DIMMs per socket we should apply following rules

If CPU is skylake

• If dimm count per socket is 5,7,9,10,11 then the model is considered unbalanced and we need to take a -50% memory adjustment
• if dimm count per socket is 2,3,4, or 12 it is balanced and memory adjustment = 0%
• if model is balanced and DIMM count per socket is 6 or 8 then it is balanced and memory adjustment = 0%
• if model is unbalanced and DIMM count per socket is 6 or 8 then it is unbalanced and memory adjustment = -50%

 

hi everyone

• We went live with our May sprint

1. Numerous updates to various vendor products – Nutanix, Dell XC, Lenovo, Cisco, HPE, Dell PowerEdge

2. Show Extent Store in Sizing Details

There has been discussion on Usable Remaining Capacity in Sizing Details compared Extent Store.

Usable remaining capacity is what is left after accommodating workloads, RF for the workload, snapshots, etc but also include savings such as compression.

Usable remaining capacity should reflect what the customer will see after deploying their workloads as you defined in Sizer.

Extent store is what storage is “available” for the customer given the recommendation. That value is simply the raw capacity less CVM. It represents what is available to them without any workloads

It is not one is right and other is wrong but answering two different questions and so we are providing both those answers
• In May sprint (now live) you see the usable capacity and now the Extent Store values in the Sizing Details.

Again usable capacity is what is left after accommodating their workloads, while Extent store is what is in totally available to them

• In June sprint we will bring over the Extent store into charts

 

Hi everyone

We launched the April sprint today. Key things in this sprint

• Sending a quote for SFDC users now includes support and licenses and not just hardware. We want to tighten the integration with SFDC and more things coming.

• Added Calm support. Now you can select if you want any workload managed by Calm. Sizer then will collect all the workloads requiring Calm and provide the proper SKUs and quantities in the BOM

• VDI Financial Analysis. Now for VDI we are providing an extensive TCO analysis

• Now support NFS in AFS’ application storage along with current SMB protocol

• Improved Storage Calculator by adding more HDD options

hi everyone

We posted our February sprint last night

This sprint we did a lot of development for the future. We are enhancing automatic sizing options and enhancing our product data models to allow for lot of product component options from multiple vendors. This is work for the near term future

What we delivered in February is important though:

1. Storage calculator like Designbrewz. Later we will replace what is in Designbrewz.com itself

2. Block Awareness. So for any workload you can have block awareness and we will put the nodes in separate blocks

Attached a couple graphics on the calculator

Hi everyone

We just posted January sprint and this is a HUGE sprint in terms of features you will love

1. First and foremost you can now clone a scenario and apply it to a different vendor. This is HUGE feature for SW only business model. Do all your hard work capturing the customer requirements in one vendor. Then you can clone that and apply to any other vendor in less than a minute. So easy to compare different vendor recommendations. Full notes in the FAQ

2. Cluster level filters. Our sizings are getting more complex. Though Sizer has supported multiple clusters for long time, most did not use it. With DR it is more common and other use cases. So now can look at All Clusters or pull in pull down just select what you want and then the Sizing details, dials, DesignBrewz, and Rack View show just that cluster.

3. List view in dashboard. Now instead of tiles you can have a list view

4. Updated VDI and RDSH/XenApp for various Windows versions and Office versions

5. We make core adjustment for Skylake memory

6. Added Fujitsu as OEM

7. Dell XC – added XC640-4 and added SED, GPU to various models

8. Product updates for Lenovo

9. Product updates for Nutanix

10. Product updates for HP, Cisco, Dell PowerEdge