<div>
What, specifically, is too slow? Essentials designed to be a relatively inexpensive domain controller and file server. It shouldn't be running applications. So k owing what is slow is the first step to troubleshooting why.
</div>

<div>
Cliff Galiher, 
we use it with as a server for an imaging program for xray, it gets the image from the xray device and all pcs in the nerwork use the software to see the images. 
Also it has a MYSQL database with very small schema with data and a visual studio application.... 
those two are the main tasks... 
 
Besides that antivirus, antimalware, some tools like visual studio environment... 
 
it is very slow when calling the windows explorer, or any other windows feature... .chrome is also very slow.... 
But as I said the task manager shows performance ok.
</div>

<div>
First guess is disk I/O is being hammered. *None* of those belong on any domain controller. And can be particularly troublesome on a box that is usually sold with essentials. 
 
I would be planning on a higher end server and virtualizing both essentials and an application VM to break off the applications into their own container or planning on a second physical server and migrating the data. 
 
Either way.youll have issues colocating as you have.
</div>

<div>
Cliff Galiher, 
Is it possible to consider the attached images a prove of a hard disk being hammered? 
<a href="https://filedb.experts-exchange.com/incoming/2019/02_w06/1411855/bbbb.jpg" target="_blank" class="file-inline" title="bbbb.jpg (256 KB)"><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512"><path d="M67.508 468.467c-58.005-58.013-58.016-151.92 0-209.943l225.011-225.04c44.643-44.645 117.279-44.645 161.92 0 44.743 44.749 44.753 117.186 0 161.944l-189.465 189.49c-31.41 31.413-82.518 31.412-113.926.001-31.479-31.482-31.49-82.453 0-113.944L311.51 110.491c4.687-4.687 12.286-4.687 16.972 0l16.967 16.971c4.685 4.686 4.685 12.283 0 16.969L184.983 304.917c-12.724 12.724-12.73 33.328 0 46.058 12.696 12.697 33.356 12.699 46.054-.001l189.465-189.489c25.987-25.989 25.994-68.06.001-94.056-25.931-25.934-68.119-25.932-94.049 0l-225.01 225.039c-39.249 39.252-39.258 102.795-.001 142.057 39.285 39.29 102.885 39.287 142.162-.028A739446.174 739446.174 0 0 1 439.497 238.49c4.686-4.687 12.282-4.684 16.969.004l16.967 16.971c4.685 4.686 4.689 12.279.004 16.965a755654.128 755654.128 0 0 0-195.881 195.996c-58.034 58.092-152.004 58.093-210.048.041z"/></svg>bbbb.jpg</a> 
<a href="https://filedb.experts-exchange.com/incoming/2019/02_w06/1411856/cccc.jpg" target="_blank" class="file-inline" title="cccc.jpg (321 KB)"><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512"><path d="M67.508 468.467c-58.005-58.013-58.016-151.92 0-209.943l225.011-225.04c44.643-44.645 117.279-44.645 161.92 0 44.743 44.749 44.753 117.186 0 161.944l-189.465 189.49c-31.41 31.413-82.518 31.412-113.926.001-31.479-31.482-31.49-82.453 0-113.944L311.51 110.491c4.687-4.687 12.286-4.687 16.972 0l16.967 16.971c4.685 4.686 4.685 12.283 0 16.969L184.983 304.917c-12.724 12.724-12.73 33.328 0 46.058 12.696 12.697 33.356 12.699 46.054-.001l189.465-189.489c25.987-25.989 25.994-68.06.001-94.056-25.931-25.934-68.119-25.932-94.049 0l-225.01 225.039c-39.249 39.252-39.258 102.795-.001 142.057 39.285 39.29 102.885 39.287 142.162-.028A739446.174 739446.174 0 0 1 439.497 238.49c4.686-4.687 12.282-4.684 16.969.004l16.967 16.971c4.685 4.686 4.689 12.279.004 16.965a755654.128 755654.128 0 0 0-195.881 195.996c-58.034 58.092-152.004 58.093-210.048.041z"/></svg>cccc.jpg</a>
</div>

bbbb.jpg

cccc.jpg

<div>
It's possible. But a bit anecdotal. You'd usually want to comment performance data over time, not just 60 second intervals. But given you are seeing some plateaus of 5 queues and 10MB/s on the C drive, with no spokes going higher, it'd be reasonable to think you're hitting the maximum that configuration can handle. If it could handle more you'd usually see bursts going higher occasionally.
</div>

<div>
What is your raid controller, RAID type, and disks used? 
I would say that you should be using SSD, but are using large and slow disks instead. 
How many users are connected?
</div>

<div>
Kevin Hsieh, 
RAID, SSD
</div>

<div>
Which RAID controller? Which RAID level?
</div>

<div>
Sorry I meant, 
No RAID, no SSD
</div>

<div>
So this is a server on a single 7.2K SATA drive? That would be your problem right there. Hopefully you didn't sell them the server.
</div>

<div>
The server was bought in the Netherlands by our customer's son, 
We are in Aruba 
We just provide the main application which has no problem. 
But we are making our extra mile here since our customer has no support from the hardware company.
</div>

<div>
So it could also be the hard disk. 
is there any command, tool, or procedure you can suggest to me? 
maybe an app that we can use to test it 
Before telling them to buy a new server
</div>

<div>
&quot;could also be the hard disk&quot; 
 
That is where disk I/O comes from.. They are effectively one and the same from a discussion standpoint for this issue. There are 3rd-parth (not free) disk benchmarking utilities out there to intentionally push thresholds, but you'd still want to set up performance counters over...say...24 hours to have a baseline. 
 
There isn't a quick test that will let you point and say &quot;see, that's your problem.&quot;
</div>

Project Manager

RGuillermo

<div>
Take a look at Windows Performance Monitor during the times where you think response is slow. &nbsp;That should give you a clue. &nbsp;Watch for CPU usage, RAM usage, Network usage, and Disk usage. &nbsp;It is a good first step toward identifying where you have a bottleneck.
</div>

<div>
The important counters would be disk latency, read latency, write latency, disk queue length. 
 
Good news is that a new server wouldn't be needed; just an SSD. I have no idea how expensive they are where you are at.
</div>

<div>
I don't won't any credit relative to this because I'm leaning toward what Cliff has contributed. 
 
You can run whatever version of 2016 but you never really stated whether this was a physical machine or VM? 
 
But then you mention chrome. &nbsp;As much as I like Chrome I don't like the fact that when I start Chrome I get 15 isolated processes consuming RAM? &nbsp;More so if you use extensions from Chrome Store. &nbsp;That would be RAM. &nbsp; 
 
You have a 7.2K drive, and stating 50% RAM and 15% CPU? &nbsp;Seems to point to Disk IO. &nbsp;Throw AV in the mix. &nbsp;Disk IO. &nbsp;If I'm building a gaming machine, not a server, I would rather have a iSCSI controller and 10K drive + RAM + CPU. &nbsp;I would not use the proposed server configuration to play my online games. &nbsp;Much less, host MySQL or image files. &nbsp; 
 
You have a server hosting MySQL but hosting potentially large image files. 
 
How are those files accessed? &nbsp;Every SMB request is just as resource intensive as MySQL. 
 
Your using Essentials so you don't have the options of Storage Spaces Direct (S2D) yet easily a benefit if you change the license. 
 
Is the cost difference worth the end user frustration. &nbsp;Absolutely not. 
 
Design/implementation where the end-user is directly impacted is perception. &nbsp;The fact that a project is technically completed, subjective. 
 
It is our job to bridge the gap between. 
 
Saving money at the expense of anyone's ability to do there job is a losing proposition. &nbsp; 
 
In the far past, network was an issue, before that CPU, RAM, bus speed to RAM. &nbsp; Today it is OS unless your hosting that OS in Flash. &nbsp; 
 
Not only do you have a single point of failure you have a single point of bottleneck. 
 
You did not mention whether it was a VM or physical server but keep in mind that the disk subsystem, today, is the most common bottleneck. &nbsp;It used to be CPU/Memory but those were easily fixed. &nbsp;Then it was bus speed. &nbsp;Then, even if you fix the disk IO using NVE or flash...The Operating System is the bottleneck. &nbsp; 
 
In this case, it would appear to Disk IO is the issue. &nbsp;This is what happens when you try to host MySQL and large image files on a single point of failure and the abridged version of 2016 Server. &nbsp;You won't solve that problem if you move the disk subsystem to Flash. &nbsp;Not to mention, you can't use S2D, SOFS. Clustering and so forth. &nbsp;Worst case, you have single point of failure, best case you have no IO. &nbsp; 
 
Every other component you install from AV to whatever, decreases IO that much further. 
 
I'm going to be blunt. &nbsp;You are asking about tools and some type of validation process to prove using this scenario is correct instead of asking how can I truly resolve this issue, your asking the wrong question. 
 
You might be able to resolve some of the issues by moving your Virtual Machine to Pure Storage? All flash? &nbsp;Eliminate the IO bottleneck. 
 
Remove 7.2k, no RAID, no controllers from the equation. 
 
You did not state the number of clients accessing the MySQL to obtain these XRay images and from what I recall, those are fairly large. &nbsp;Are those being served up via NTFS or MySQL? &nbsp;That makes a really big difference. &nbsp; 
 
When using S2D? ReFS versus NTFS? 
 
What happens when that server goes down? &nbsp;What happens? &nbsp;If it cannot handle the IO now...It will go offline, matter of time. 
 
It sounds like your wanting to make something work where a decision has already been made to use a specific OS that is not designed to handle your requirements. &nbsp;Whether this is coming from you or someone else, doesn't matter. 
 
It's a bad design. &nbsp;It's simply not going to work. 
 
Like buying decaffeinated coffee. &nbsp;Slightly lesser cost, none of the benefits. &nbsp;Perhaps some people drink coffee for the taste, most drink it for the caffeine. &nbsp;Your going against the grain. &nbsp; 
 
It almost seems as if someone in leadership is trying to control cost at the expense of actual functionality because they simply have no clue otherwise. 
 
Well, they are wrong. &nbsp; 
 
The cost difference for a working solution and peace of mind is almost priceless when you take the time to consider all the Help Desk tickets, lost data, no DR, or something as simple as a single server trying to host MySQL and Xray files and where that is not the intended purpose. &nbsp;Yet, what's the harm in trying? &nbsp; 
 
No harm in trying but there is no amount of &quot;tools&quot; that you can throw at this solution to make it do what it was never intended to do....... 
 
Sorry to be so blunt. &nbsp;No apologies on the brutal honesty.
</div>

<div>
This is MySQL and we are dealing with large Xray files that are sent to the server, no indication of how often and the size of the schema has no relevance. &nbsp;No RAID controller. &nbsp;7.2k drive, one. &nbsp;Single point of failure. &nbsp;If the disk is dying, bad design. &nbsp; 
 
Physical disks wear out. &nbsp;Minimum, simple, mirrored drive with RAID and cold spare for swap out solves that issue. &nbsp;0 or 1 goes bad, swap out the cold spare to rebuild the mirror, problem solved. &nbsp;Lose a single 7.2K drive with no RAID, no redundancy, assuming you have a good recent backup that compensates for every single persons XRay written....still have to replace the drive, then OS, then Data. 
 
If the drive is dying, how do you replace it without taking down the entire system? 
 
This is a fundamental design issue. 
 
What about the MySQL database? &nbsp;Is that backed up too? 
 
If not using SQL with minimum BAG.... host the DB and log files on S2D. 
 
Worse case, still have to rebuild the server....from scratch. &nbsp; Install MySQL, restore the Databases. 
 
If I come in for an Xray. &nbsp;What happens? 
 
Where does that file go? &nbsp; 
 
As a former HIPAA Officer, I want to know. 
 
Why take this risk? &nbsp; Am I missing something? 
 
Seeing the Forest...Despite the trees.
</div>

<div>
Single point of failure is one thing. &nbsp;Single point of failure with medical information that is specific to a patient and in most cases that information is needed on demand. &nbsp;Times number of patients getting Xrays per day written to a MySQL instance on a 2016 Essentials Server with no Raid and 1 physical 7.2K drive. &nbsp;Assuming it is a physical server your looking at a full restore, best case full metal, after replacing the drive and depending on your hardware support agreement this can be anywhere to 4 hours or 48. &nbsp;Unless you have drives in inventory. &nbsp;Assuming you have a backup, restore, is MySQL fully backed up. &nbsp;Every transaction, every write. &nbsp;How many Xrays have you lost? &nbsp; 
 
Seems rather critical to me. &nbsp;What exactly am I missing here? 
 
I can only compare this to the # of Hospitals I've been in relative to Citrix migrations, Epic, Hyperspace, and other technologies where a person's Xray or MRI was critical to diagnosis. &nbsp; 
 
The overall design is wrong. &nbsp;No HA, no mention of backups. &nbsp;Single Drives, No Raid. &nbsp; 
 
Is this an IT problem or a doctor office misunderstanding of the requirements relative to IT and the necessary cost to implement the right technology? 
 
Is someone being &quot;constrained&quot; from a lack of understanding? 
 
Let's assume this is a doctors office where Xrays are used. &nbsp;They have an Xray machine. &nbsp;Doctors, understand that cost. 
 
What they seem to miss IMO is the fact that those XRays must be stored somewhere. &nbsp;Let's use MySQL instead of SQL. &nbsp;Perhaps they are fine with having x# of patients &nbsp;coming back to the office to inundate themselves with more radiation to get more images because my single single with a single drive failed. &nbsp;Do they not care if the backend goes down or is their a miscommunication? &nbsp;Are they even looking at the Xrays? &nbsp;Is this a Chiropractic office? &nbsp; 
 
Please, put me in contact with that Doctor. 
 
I'm getting the impression, the Xrays, don't matter so much. 
 
And that is a problem for me.
</div>

<div>
The blue line on the disk Graph is showing it’s running out of IOPS! 
Might only be doing 10MB/s but it’s showing a steady 95-99% IOPS useage! 
That would slow things down!
</div>

<div>
Assume the majority agrees you have an IO issue. &nbsp;You have a single server, no RAID, one Drive, 7.2K. 
 
This is obviously the issue, how do you resolve it? 
 
Well, you can switch from MySQL to MS SQL cluster with BAG. &nbsp;You need two more servers. &nbsp;In return, you get a clustered SQL (Standard). &nbsp; 
 
You mentioned a small Schema. &nbsp;Where is the Xray data actually be written? &nbsp;Is it MySQL or a SMB share with references to this data? 
 
Do you have a flash array? VMWare? Can you port this server to a VMWare type hypervisor with multiple hosts and failover at will? 
 
IOPS is the least of your concern. 
 
This is real world, real people, medical information being stored on a single server with no RAID and a 7.2K drive. 
 
If the Xray machine takes an image and cannot write it, what are the downsides? 
 
Does the software support MS SQL or only MySQL? &nbsp;MySQL is fine I suppose if I'm not storing patient data or Xrays or MRI's. 
 
Might as well host it on Windows 7 Professional. &nbsp; Or maybe 98 with an iSCSI controller and a 7.2K drive. &nbsp;One of my best gaming machines many years ago. 
 
You simply cannot store patient data on a single server with a single drive running Windows with MySQL and zero redundancy. 
 
So what if the drive is failing. &nbsp;Throw in a 10K drive. &nbsp; Rebuild the entire system being you have no RAID, no Mirror, now way to swap the drives and not destroy all the data. 
 
Then what? &nbsp; You went from 7.2K to 10K? &nbsp; 
 
I'll throw the towel in anytime you want but for some reason I see a much bigger issue here. 
 
You cannot solve an IOPS issue with comments you have an IOPS issue and no resolution. 
 
You have to get away from Essentials. &nbsp;If high end storage is not an option you at least have S2D and surely the application supports SQL (Standard Edition) so you can minimum configure a SQL cluster and if the images use SMB, you have the S2D with SMB 3.1. &nbsp; This is cheap compared to any form of enterprise storage. &nbsp; 
 
This can be virtual or physical. 4 physical servers and Hyper-V. &nbsp;Spin up guests for SQL, S2D, and leverage the local storage across all hosts. &nbsp; Full redundancy and localized HA. 
 
We can discuss the fact that you have an IOPS issue. &nbsp;Or we can discuss how to solve the fundamental design problem.
</div>

<div>
(Kevin) &quot;Do you have a flash array? VMWare? Can you port this server to a VMWare type hypervisor with multiple hosts and failover at will?&quot; - this was rhetorical not a recommendation. 
 
One of multiple examples, recently implemented Storage Spaces Direct (S2D) at a major hospital system (Methodist) with SQL 2016 Cluster and Basic Availability groups across two active-active Data Centers and GSLB on Netscalers and 12,000 users. &nbsp;Not to mention, it was signed off on by the entire Infrastructure Committee and having assessed and test 7 other options. 
 
S2D hosting user profiles for 12,000 users. Active-Active data centers with GSLB hosting the most critical application in the organization, Epic, and that was just to get at the medical/patient data. 
 
For this scenario mentioned above - having two servers, versus one, and local storage with SD2 where the technology is highly optimized with caching, mirroring and in combination parity would allow you to build out a complete solution with a few physical servers and the ability to obliterate any performance numbers relative to IO had at present with the single server, single failure solution, no HA. 
 
Doesn't matter the size of the &quot;clinic&quot;. &nbsp;Once you decide to start accepting people in a medical capacity and storing medical data that data must be secure, highly available, and has retention requirements. &nbsp;What is the vendor reference architect design suggestions? 
 
With ReFS and S2D you get real-time tier optimization, block cloning, and sparse VDL and it is built in to 2016/2019 Server and have multiple options on the drives because the IO is spread across. 
 
Storage Spaces Direct uses local-attached drives to create highly available, high scale software-defined storage at a fraction of the cost of traditional SAN or NAS and fits in converged or hyper-converged scenarios. 
 
Storage Spaces Direct is defined as &quot;software-defined shared nothing storage.&quot; It creates a pool of storage using the disks in a cluster, with new disks getting automatically added to the pool. Organizations can benefit from Storage Spaces Direct by tapping a collection of less expensive storage devices, such as NVMe, SATA and SDD. 
<a href="https://sec.ch9.ms/ch9/f960/f12f08d4-8cb6-4bdc-8765-cbca789af960/StorageSpacesDirectDesignConsiderations_high.mp4" rel="ugc">https://sec.ch9.ms/ch9/f960/f12f08d4-8cb6-4bdc-8765-cbca789af960/StorageSpacesDirectDesignConsiderations_high.mp4</a>&nbsp; 
 
The comment above... &quot;That's why we have SSD, and now NVMe drives.&quot; &nbsp;- That combination is is part of the S2D design options. &nbsp; 
 
Yet, Adding S2D or NVMe to a single point of failure doesn't solve all the other issues relative to HA and data retention for medical data. &nbsp;My point, still single point of failure and won't address latency issues. &nbsp; 
 
Doesn't matter how small the &quot;office&quot;. &nbsp;This is patient data. &nbsp;NVMe drives in a single server turns the server OS to bottleneck. &nbsp;S2D with caching and other technologies helps buffer and eliminate those top of stack issues while at the same time providing full redundancy. &nbsp; 
 
Storage spaces is not going away and will only get better. &nbsp;Not to mention, Intel is involved. And, it is not all about IO, what about latency? 
 
A new record was set recently with 2019 and S2D 
With 100% reads, the cluster delivers 13,798,674 IOPS. 
With 90% reads and 10% writes, the cluster delivers 9,459,587 IOPS. 
 
Conclusion: 
Together, Storage Spaces Direct in Windows Server 2019 and Intel® Optane™ DC persistent memory deliver breakthrough performance. This industry-leading HCI benchmark of over 13.7M IOPS, with predictable and extremely low latency, is more than double our previous industry-leading benchmark of 6.7M IOPS. What’s more, this time we needed just 12 server nodes, 25% fewer than two years ago. 
 
Let's discuss Hyper-V and CSV 
Storage Spaces Direct is compatible with the Cluster Shared Volume (CSV) in-memory read cache. Using system memory to cache reads can significantly improve performance for applications like Hyper-V, which uses unbuffered I/O to access VHD or VHDX files. Unbuffered IOs are any operations that are not cached by the Windows Cache Manager. 
 
Because the in-memory cache is server-local, it improves data locality for hyper-converged Storage Spaces Direct deployments: recent reads are cached in memory on the same host where the virtual machine is running, reducing how often reads go over the network. This results in lower latency and better storage performance. 
 
<a href="https://blogs.technet.microsoft.com/filecab/2018/03/27/storage-spaces-direct-momentum/" rel="ugc">https://blogs.technet.microsoft.com/filecab/2018/03/27/storage-spaces-direct-momentum/</a> 
We’ve reached an exciting milestone: there are now over 10,000 clusters worldwide running Storage Spaces Direct! Organizations of all sizes, from small businesses deploying just two nodes, to large enterprises and governments deploying hundreds of nodes, depend on Windows Server and Storage Spaces Direct for their critical applications and infrastructure. 
 
The size of the organization doesn't matter. &nbsp;When no one can access their Xrays, the finger will be pointed at you. 
 
Addressing these issues is a &quot;cost of doing business&quot;. &nbsp;And keeping it. 
 
You can buy another hard drive &nbsp;but it won't solve your immediate or long term problems. &nbsp; 
 
It is simply easy to state you have an IO or performance problem but then not propose a real option to resolve it. &nbsp;This or that monitoring tool that proves you have an IO problem is not a solution. &nbsp;It's just someones way of validation you have an IO problem. &nbsp;I'm simply proposing a viable cost effective solution to the holistic problem. 
 
What about all the other issues relative to design, patient data, single drive single point of failure, single server point of failure, backups, and data retention, and DR. &nbsp;Your hosting Xrays of patients. Medical information. 
 
Most of this can be solved with two servers, as stated above, clustered with S2D, ReFS which is native to 2016 Server at no additional cost. &nbsp;Then, licensing for Microsoft SQL which will run on the same cluster and HA with Basic Available Groups. Two servers versus one, the right Operating System and Microsoft SQL instead of MySQL. &nbsp;Otherwise, switch to Linux and run MySQL if your only goal is to save money. &nbsp; 
 
Something is going to happen and the refusal to do something right and spend a few more dollars now is going to cost you hundreds of thousand of dollars in fines and losing patient data resulting in the &quot;clinic&quot; going out of business. &nbsp;Avoided for the cost of two servers running a higher version of Server 2016. &nbsp; 
 
Seeing the forest despite the trees. 
 
That server and version of OS, cannot handle the load and when it dies you have no recourse. Plan ahead, versus fire drill mode, and have two servers and spread the load using native embedded technologies in 2016/2019 and use SQL 2016 with BAG and clustering, Storage Spaces Direct, for a very small cost relative to losing everyone's medical data and address this from a holistic perspective and address your IO issue, HA, and now that you have those basic requirements start contemplating a backup strategy with RTO and RPO. &nbsp; 
 
That medical data has high retention requirements. &nbsp;May need access to it a year from now or 20 years from now or it could be 5 years. &nbsp;That retention needs to be defined and signed. &nbsp; 
 
What about the vendor of the medical equipment? &nbsp;What is their proposed reference architecture for hosting the patient data? &nbsp;I highly doubt it is Essentials and single server with single drive and single point of failure and no data retention strategy.
</div>

<div>
I just sent all this discussion to the company owner. 
They gave it to their Local It company or technician I guess. 
 
Some information got to know and I can share is 
there is no Virtual machine. 
They do have a backup software working every 
&nbsp;six 6hours as they inform me. 
No loss of xrays yet 
 
But I did make clear the warnings the OS, it's not appropriate 
for those apps. 
 
I am seeing they will react pretty soon in the next few days 
attending your indications. 
 
Thank you to all Experts for your excellent advice.
</div>

<div>
I just don't see how an organization that just bought a server with a single spindle is going to go ahead and spend the money on Microsoft Windows Datacenter because it will save money on fancy storage. 
 
Two servers, with appropriate storage, Microsoft Windows Server Standard, and Veeam will cost less than a single Datacenter license. I bet you could throw in some cloud storage for offsite backup and recovery too.
</div>

<div>
<div class="content wysiwyg-content">
Hellos Experts, 
Our customer has a WINDOWS server 2016 Essentials 
it is simply too slow...even unresponsive many times in a day. 
 
Task manager reports 50% of ram used and less than 15% of CPU usage. 
 
but still slow overall... 
 
What can I do in this case? 
Regards,
</div>
</div>

Hellos Experts,
Our customer has a WINDOWS server 2016 Essentials
it is simply too slow...even unresponsive many times in a day.

Task manager reports 50% of ram used and less than 15% of CPU usage.

but still slow overall...

What can I do in this case?
Regards,

Win server 2016 too slow

This topic area includes legacy versions of Windows prior to Windows 2000: Windows 3/3.1, Windows 95 and Windows 98, plus any other Windows-related versions including Windows Mobile.

Windows OS

Windows 10 is a personal computer operating system featuring the "universal application architecture" (UAP); apps can be designed to run across multiple devices with nearly identical code, including PCs, tablets, smartphones, embedded systems, Xbox One, Surface Hub and HoloLens. Windows 10 also includes a virtual desktop system, a window and desktop management feature called Task View, the Microsoft Edge web browser, support for fingerprint and face recognition login, voice-based search (Cortana), new security features for enterprise environments, and DirectX 12 and WDDM 2.0 to improve the operating system's graphics capabilities for games.

Windows 10

Microsoft Azure is a cloud computing platform and infrastructure for building, deploying and managing applications and services through datacenters. It provides both platform-as-a-service (PaaS) and infrastructure-as-a-service (IaaS) services and supports many different programming languages, tools and frameworks, including both Microsoft-specific and third-party software and systems. Cloud Services is a PaaS environment and can be used to create scalable applications and services; there are specific software development kits (SDKs) provided by Microsoft for Python, Java, Node.js and .NET. Azure also has file and storage services, data management, analytics and DNS services.

Azure

Windows Server 2016 is the successor to Windows Server 2012 R2. Built upon the same core code as Windows 10, Windows Server 2016 brings enhancements in security, servicing, and connectivity. A particular focus on this release was hybrid-cloud scenarios, and has close ties to Azure and other Microsoft cloud initiatives. This does not detract from the many improvements that are available for on-premises-only deployments

Windows Server 2016 comes in Datacenter, Standard, and Essentials editions, and for servicing, has adopted windows 10's cumulative model. The new nano-server install is designed to be remotely managed and is designed to be kept current through continuous feature updates. The full GUI install operates similarly to windows 10's "Long Term Servicing Branch" (LTSB) model with cumulative security updates.

Windows Server 2016 has also shifted from a per-processor-and-CAL licensing model to a per-core-and CAL licensing model. This brings Windows Server's licensing more in line with Microsoft's other products and makes hybrid-cloud license planning easier as well.