What StarWind offers to solve uptime problem for SMB and ROBO
As you might know, the small and medium businesses and ROBO really have a problem with IT departments, that is either it’s a one-man shop, and they simply don’t have dedicated IT personnel or its ROBO, or they may have one IT guy for ten locations, which is also not the best idea. Since for small businesses it is critical to provide best level of service for ROBO, usually those are the locations where you face the customer, so even if you’re a big car-rental chain, all those small airport locations are your face, and if the guy in the small location says “I’m sorry, but I cannot get you the car because our system is down”, – that is where the negative experience starts from.
No one may notice a big outage in your main datacenter, but if something happens on the edge where you meet the customer, that’s going to have dramatic circumstances. So, StarWind idea is to take a solution and bring it to potential customers which are SMB and ROBO, and the idea is to solve the problem of IT complexity. And our solution to IT complexity is HyperConverged Appliance, which unifies all the pieces of the puzzle together and doesn’t ship you a box with jigsaw puzzle parts, but it ships you a ready puzzle which is already glued to a cardboard. It’s already beautiful, you can use it, and you know how to use it. If something happens, there is StarWind support for the entire structure. That is how we take that complexity out of the game and make sure that IT is easy for any environment sites.
We recently expanded our line up and previously it was just small, large and extra-large lines in StarWind Line Up; right now, we have also added the XL Ultra, which is really high spec hardware, and the main idea remains the same: you get two servers, and two servers provide a highly available resilient IT virtualization environment for your business. And these solutions are fit for any task; they can be scaled-out or scaled-up, so these are just the basic building blocks.
So, what makes StarWind completely different from all other hyperconverged solutions on the market? First of all, as you may have noticed, we start with two nodes. It’s a minimum possible scenario for a highly available infrastructure, and we don’t need the third node for voting or for keeping additional setup, you can start with two. If your environment grows, you can scale it to three nodes, to four nodes, but to start with we don’t need that big of a jump.
Your infrastructure can be modified and grown with no extra hardware expenses
There are multiple implementations of StarWind HCA
. Let’s say, we have a customer who has multiple nodes in their head office; they have two XL systems running their VDI deployment, they have three L systems running their virtualization, and they are doing replication between the sites, they have, actually, three branch offices drawn, and technically ten of them are doing replication between head office and ROBO locations. So, that is one of a typical use cases for StarWind HyperConverged Appliance. We’ve already mentioned that, but didn’t go into details yet – so, what makes us different: first of all, the two-nodes setup, which is one of the few offerings on the market, and next thing is that we can modify our appliances and we don’t prevent our customers from doing so.
So, if you want, there is a standard configuration. You get on your purchase from StarWind, and then, with any hyperconverged vendor if you want to grow your infrastructure, you need to buy an extra server, and with StarWind it’s not like that, it’s much easier: you want to increase the CPU horse power – okay, put in the new CPU, we do support it. If you want more RAM or more storage, putting more drives and more demo sticks, we do support it. If you want to scale out, that can be easily done: you’ve got two servers upgraded to four, five, six or up to 64 servers, however you like. Then, if you want to scale up, add storage to individual servers or to old servers, StarWind will still support that. So, this is a really unique offering and it is something, which allows you to really tailor the solution you’re building for your IT environment.
So, we talked about the surface, we talked about the hardware footprint, and now let’s look under the covers. What makes it possible to use just two servers and get a highly available server virtualization environment? The whole background synchronization clustering is happened on the hypervisor level where StarWind does full storage mirroring between the servers providing a highly reliable and highly available storage resource for your virtual machines. Then, there is Microsoft Hyper-V failover cluster providing virtual machine clustering and using the StarWind storage to have a fault-tolerant virtual machine repository; and then there is 5nine, which is used as a single pane of glass management for the entire architecture, and then there is Veeam because there are no good infrastructures without a backup. Any good infrastructure needs the backup and that’s an absolutely mandatory requirement. Even if you have one node running two virtual machines, it’s always good to have your backup.
With StarWind working with local RAMs and disks is faster
Now, you may be wondering why didn’t we go with the standard approach where people buy a SAN, they buy multiple servers. So, basically all that stuff you hear from vendors, when you go to their website and try to quote a system or try to build a system – they say you need a huge fiber channel SAN, they say you need a blade channel, they say you need multiple servers to do that. But here is the thing – with a traditional SAN the I/O
from the virtual machine running your mission-critical applications first goes to the hypervisor, then to initiator, physical or virtual one, then through the storage network, and only then hits the cache on the storage solution, and only after goes to disk.
With StarWind, since we are hyperconverged and we omit the entire SAN layer, first the virtual machine writes data down to hypervisor, and then StarWind intersects that and puts in directly into server’s cache, which is DRAM, the fastest memory resource we have in the servers, and then from DRAM that data travels down to disk and we keep it avid and block-level synchronized, and that includes the cache as well; so, if one node goes down, the data will still be available on the other node, in cache and on spindles, and on flash, of course, so that is how we maintain everything highly available.
Joint solution enables management of the whole architecture with no overkill
Now, thing peculiar to StarWind HyperСonverged Appliance
is a collaboration between StarWind, Veeam and 5nine to deliver a joint solution, which really stands out and allows customers to ease the management of the whole architecture because typical solutions like System Center may be an overkill for a small environment, they may be an overkill even for a big environment, so we decided we need an easy and good-working alternative, and with 5nine and Veeam we found that alternative. We really have an excellent management console which allows to drive the entire cluster all the way from a single virtual machine up to virtual networking, best practices analyzing and much more interesting features. All pieces of the puzzle are managed with 5nine Manager console and this includes all layers, starting from storage and up to the virtual machine.
Now, it is also possible to manage the StarWind storage through 5nine Manager console, monitor the performance, see the cache utilization deduplication ratios and stuff like that. It’s all available here, just on the other tab, and of course the integration was the backup solution, so you just switch to the VM tab in the 5nine Manager and you can see your backup jobs, you can see what was wrong with backup jobs, which jobs completed, which did not, and you can control it all from here – really useful feature for an environment.
So, yes, as I mentioned, all the necessary information is now up and this is really neat and you don’t need to go through multiple consoles and navigate and filter the events to see if anything was wrong, you see it all in one console, and there is a summary showing you that your infrastructure is okay, everything’s up and running, all the backups are fine and we’re good to go.
Now, let’s talk about 5nine Manager. The company was found in 2009 right when Hyper-V
was first introduced, and 5nine has the only available solutions from Hyper-V. This means that 5nine has some competitive advantages. When they integrated at a very low level they worked extremely closely with the engineering team and set quite a lot of performance optimizations which no one else in the industry is doing either for management or for security. 5nine is based in the United States, has now staff in 25 cities worldwide, and has 18 Microsoft MVPs as part of team. They have about 80 000 total Hyper-V users across all the different products. In addition to 5nine Manager, there is an agentless security solution which basically means that you can protect your entire virtual infrastructure without ever having to install anything inside a virtual machine. 5nine intercepts all of the security packets and network traffic at the host level.
Additionally, if you are currently a VMware customer looking to move over to Hyper-V, VMware to Hyper-V immigration is offered for free, as well, so that you may be interested in checking out. As mentioned before, the company is based in Chicago, but it is expanding relatively fast worldwide with a lot of focus over the past year in expanding throughout Europe and Southeast Asia, and the big push throughout 2016 is going to be continuing to expand throughout Asia, as well as the Latin American market.
The whole goal around 5nine Manager was to provide an easier way to go in managing Hyper-V infrastructure. For those of you that are using the inbox tools you probably found that it could be kind of difficult to learn or understand all the different places that you have to set up a cluster. In some cases, it is six different consoles you have to go to get a cluster up for running. So, the intention of 5nine Manager is to make it easier for people to learn Hyper-V, to get familiar with it. And with this they found a lot of success with VMware administrators that are using Hyper-V for the first time because the console actually looks like VMware vCenter, so it makes the transition easier and makes the evaluation process much smoother if you’re looking to move. Also, they’ve taken many of the features of System Center Virtual Machine Manager that SMB want to use.
A lot of customers out there really want to use System Center to have the centralized management, but are unable to because System Center itself can be pretty expensive. It includes a lot of additional components and actually requires dedicated hardware just to go on running management system. So, 5nine Manager is trying to address all of those problems with having a unified lightweight management solution for all of your virtualized resources.
The first thing that all lab men should be doing every morning is going to see a Summary View. It is the place where you can roll up information across the entire clusters, hosts and virtual machines, so they could quickly identify problems and view any types of alarms that come through the different channels.
An easy way to manage your Hyper-V
host is provided. The actual first implementation of 5nine Manager is a full graphical user interface which runs directly on Windows Server
core and Hyper-V Server. When these versions were first introduced in 2008, it was really hard to do remote management, so people struggled deploying that.Even though these versions are supposed to have zero graphical user interface, 5nine Manager itself can actually run directly on these boxes giving you a full GUI, so that you could do all the configurations locally without having to learn PowerShell or having any other dependency on other scripting languages. They never overextended that giving you the ability to do direct management or a lot of remote management, including down level management all the way back to 2008 r2, which is something that even Hyper-V Manager doesn’t support. It’s easy to add new host automatic discovery through AD, and once you configured one of your hosts perfectly, 5nine gives you the ability to clone the settings of that host to all of the other hosts in your environment.
Virtual machine management
5nine Manager also supports every version of every Hyper-V virtual machine, so it doesn’t matter whether you’re running Windows, Windows Server, Linux or UNIX, 5nine is going to be able to support it, and that even includes the next versions with Windows Server 2016, along with Nano Server.
Virtual resource monitoring
There is basic resource monitoring, as well, so you can take a look at current as well as historical indicators for the CPU, memory, network, and disk i/o having the key indicators about the health of your virtualized environment. And you have the ability view this at the cluster level, host level, or even down to the virtual machine level, so you really can go and isolate the root cause of a problem whenever something happens.
Also, there is an option to view historical data. Let’s say, you come to work on Monday and something is crashed. Well, it’s now easy to just go and take a look at the performance trends over the past several days and you can see the exact time when that problem occurred, so this is going to save your money just by simplifying the troubleshooting process as well. And in one of the most recent releases even capacity planning has been added, so this is going actually help you when you are expected to run out of disk space. And for smaller organizations, this can be important, because you want to really try to defer cost as long as possible, don’t have to go and spend that extra couple thousand dollars on the new rack too soon. And without capacity planning tools, it can give you a great guidance about when you may be running out of disk, CPU or memory, so that you know in advance when you can call StarWind and order your next direct at the appropriate time.
5nine wants to make management easier as well for people that don’t have a full troubleshooting team. So they integrated all of the different blocks for your entire infrastructure over into a single console, so no need to go through these different eventing channels and different locations to see the health of your infrastructure.
Email alerts and notifications
Also, as far as for many SMEs it is not realistic to have somebody sitting in front of the console just watching it for problems to happen, 5nine have integrated email alerts notifications. Whenever something unusual or unexpected happens in the environment, you should go triggering e-mail notification, it’ll send you an alert saying: “Hey, something’s crashed and your VMs fail to deploy, something’s replicated or removed to a different host”. This means that you can now be reactive to the problems as well, saving your time instead of having to login regularly to go and check the health of your infrastructure.
Best practice analyzers
One of the greatest features of 5nine Manager, which unfortunately a lot of people didn’t get out of the basic Hyper-V product is the integration of all the different best practices analyzers. These are actually several different best practice analyzers built into Windows Server
with a hidden in different places: spread out across Server manager, Hyper-V manager, different PowerShell Cmdlets and Failover Cluster
Manager. There was an intention to unify them and put them all into a single console and a single view, so that you can get these best practices to really help you optimize and provide resiliency for your infrastructure. The recommendation to customers is to run this once a week; it will give you a report, show you changes in the environment and try to solve each and every one of these, keep running it again till you start to get a fully optimal Hyper-V environment. This way you can have the highest availability for all of your service and the least amount of disruption to any type of downtown.
Built-in system reports have also been introduced. This is going to allow you to generate a pdf, or excel, or word format showing all the different configuration components in your infrastructure.
If you work in any type of the industry that requires compliance, document into the configuration of your infrastructures is a key requirement. This includes financial, health, legal. All of these organizations that maintain certain sensitive information, have these different compliance requirements. We make it extremely easy to document the health of the environment, to adjust configuration settings down to the individual variables for every virtual machine. You can see which of them are prioritized against each other.
Failover cluster management
Let’s take a look at some of the more advanced features. We talked a little bit about cluster management already, but, again, we found that most virtual machines need to be highly available, so we want to make failover clustering a first-class citizen, giving it the ability to create clusters, do health checks, review the cluster health, move cluster nodes and, of course, traditional live migration.
Hyper-V Replica support
5nine Manager even supports Hyper-V Replica something which System Center doesn’t support, and Hyper-V Replica gives you the ability to take a virtual hard disk and replicate out to a different location as a disaster recovery
solution. All of this is integrated right within the 5nine Manager console. 5nine gives you the ability to load-balance VMs between different hosts; within System Center this is known as dynamic optimization and, say, thresholds to actually go and move resources around when they’re exceeded.
Change tracking and user logging
5nine provides you with a resource and template library, so if you’re going to redeploy the same type of virtual machine multiple times, you could basically create a golden copy of it, the information is unique and you go redeploy it as many times as you need giving you a quick and easy way to scale-up while avoiding the chance of misconfiguration, if you manually add it in this each time. As have been planned, we’ve also integrated change tracking and user locking, so any time you use locks-in, it makes change to the environment, or any time that the system or 5nine Manager makes a change, such as during a load-balancing operation, 5nine will constantly inventory this operation, so you can always go back and figure out who’s made what change and when it happened.
5nine Manager with Antivirus
There is also an advanced version of 5nine Manager that includes agentless antivirus. The target here is essentially take a logic from the security product, that gives you the ability to learn on how to install an agency on every VM integrated this with 5nine Manager. This means, you have an option of using antivirus from the Bitdefender, Kaspersky Lab or ThreatTrack and 5nine basically takes these antivirus engines and optimizes them for Hyper-V.
What this means is as this network packing actually goes through the Hyper-V host to the virtual switch, to the VM itself, it is intercepted in network packing and actually inspected, 5nine team checks whether it’s safe, makes sure that it does not contain any types of viruses or any other type of malware and the way to do this is Hyper-V itself has what’s known as an extensible switch. So, this means that security vendors or network vendors, could unplug an agent derailed into the networking stack that allows to see all the traffic is going north and south to say whether it’s safe on it or to say: “Hey, this has some type of threat, it shouldn’t be allowed through”.
There is not only protection against North and South traffic going from the host to the VM’s, but also protection against East-West traffic where you actually update going from one virtual machine to another virtual machine, giving you the ability to really have entire network protection for antivirus.
Hide security with Agentless Protection
So, why would you want to have a solution that does not require agents inside the VM? First, if you have end-users that are non-technical, it’s much easier to manage essentially. A user never sees security, they never have to update definitions, they could never disable it, whether that’s purposely or accidentally, and so they realistically don’t even know that they’re being protected. Now, if you have compliance needs, as well, that sets a lot of competitive advantage. Since you know long as an administrator, need to run a script to install the agent, it gets even more protection for that VM. You can expect everything of the host level, never needing to login, so you need increased privacy requirement. You simplify VDI management, you can just manage everything in a single location.
AV detection on the network
What 5nine essentially does here is as the strike comes through, they’re mirroring it, open up network packet and are actually doing a scan on the data, and if a threat is detected, that is actually within the network packet, it won’t be allowed to go through, it’ll be blocked or there’ll be an alert saying: “Hey, this virtual machine has just downloaded a bad file, you need to go and isolate a file”.
AV security on the disk
Once that virtual machine takes that file and still is going to write on the disk at some point. So, one of our big patents is the agentless protection. The second patent that pending is to have the industry’s fastest antivirus scans. When you do an antivirus scan it takes up a lot of CPU and memory known as a scanning storm. This means that VMs slow down and you can have fewer VMs per host, but 5nine implemented its own patent-pending technology, it doesn’t require you to scan the entire disk each and every time. All what’s supposed to be done is to scan the blocks of data that have changed. So instead of one terabyte, maybe we only one megabyte or a hundred megabytes have to be scanned. And this efficiency in the benchmark test is going out to seventy times faster than some competitors, so this means that as end-user, you don’t get hit by antivirus scans so that you can run more virtual machines on each and every host, ultimately saving you money and get a better density.
Antivirus built and optimized for Hyper-V
This has been again built exclusively for Hyper-V
, so Microsoft certified and you get a choice of three different antivirus engines, and similar to the support policy, which StarWind has here, everything rolls up into a single vendor. 5nine handles all of the licensing with Kaspersky, Bitdefender, and ThreatTrack on the backend. You just select which works best for you.