We compare what is impossible: cheap hosting and cloud on the VMware stack
The reason why we decided to write this post is due to the existing misunderstanding of the significant differences between hosting discounter services and corporate cloud provider services. Of course, a logical question follows from this: “Why do cloud services cost more?”, Because prices are often the first to be compared. The cloud server “loses” to the conditional hosting “for 99 rubles per month” for the price, but after all, it is necessary to compare “apples with apples”, and not “apples with a tangerine slice”. It is good that you can try this and that, eventually realizing that it suits “to taste”.
Hosting as a business is relatively simple. There are several options for the creation and development. You can become a reseller of a large hosting company or rent a dedicated server, or purchase your own equipment and place it in a data center. You will also need to select, install, and configure a web hosting control panel that allows you to manage software, such as a web server, database server, mail server, and other related programs through a web interface. It remains to solve the problem of technical support, start marketing and find customers. And it seems that nothing special, but you need to meet the competitive price, conditional "99 rubles per month", and therefore have to save.
')
Loukosterov use inexpensive and often problematic equipment. How to buy an expensive server and storage, but at the same time meet the competitive price? On Habrahabr, you can read about the case of failure by a major provider in the request to replace the old disk with a new one and install it under the guise of a new disk of the same old one. Disks “fall off” and you lose data. Everyone can independently assess the value of their corporate data, as a rule, it exceeds the amount of savings by at least an order of magnitude.
Partially solve the problem can backup. But this service is provided not by all cheap hosting, and if it offers to make copies of data, then perhaps it will be the same disk on which your working data is stored. In the event of a disk crash, your data and backup files are lost.
The second important component in the work of hosting is technical support. A 24-hour support service will require the hiring of qualified staff and a substantial payroll. What if the owner of the hosting decides that he will support the operation of the hosting on his own? 24 * 7 mode is canceled, because you also need to sleep. Problems happen at the wrong moment, customers are forced to wait and suffer losses.
The choice of a hosting provider in this market segment is complicated by the lack of a free trial period. The client is not able to first try the service, and then pay. Maybe someone will be lucky and there will be no problems with hosting, but first they have to take risks, since the low-cost airlines ask for 100% prepayment. In case of failure, money back programs will partially cover monetary losses, but not wasted time.
The business has needs that many hosting discounters cannot meet. When choosing a hosting you need to check
So, cheap hosting is the lack of guarantees. This is how economic systems are structured, which is all due to something or someone else’s expense. If customers pay only "99 rubles per month", there is no possibility in the framework of commercial activities to provide services for a large amount. So you have to save on hardware, technical support and limited to a narrow functionality of the software.
There are projects for which this service will prove to be an ideal ratio of price and needs to be satisfied. However, if greater reliability, safety and flexibility is required, there are cloud products that, with a significantly higher price, have a higher customer value / price ratio. Comparing such services should, of course, be compared with similar services, and not with cheap hosting.
To better understand how the business of a cloud provider is structured, let's start by analyzing the concept of SDDC.
If we translate Software Defined into Russian, we will get the term "software-defined" or "software-configurable." However, the trend of shifting the focus from “hardware” to “software” better describes the first option, therefore we will use it as the main one.
So, Software Defined Data Center or SDDC is a software-defined data center. This concept was first proposed by Stephen Herrod in 2012, who was at the time the position of VMware Technical Director.
Such a late emergence of the SDDC concept is associated with the processes taking place in the IT industry. In the 1950s, the cost of hardware was disproportionately higher than the cost of programs, which were considered mere additions to hardware. That is, the first task was to develop a hardware solution, and a secondary one - software.
Gradually, the process of komoditization or the transition to the creation of more universal and interchangeable computers led to the understanding that the program can be used on any computer equipment. As a result, by the age of 80, software development becomes an independent industry, and the cost of software first approaches the cost of developing hardware, and then in some cases begins to exceed. This overrides the primacy of hardware in favor of software. All the specialized functionality is transferred to the software, and the hardware becomes more versatile. The process, extending to all elements of the data center system, creates prerequisites for the creation of software-defined data centers or SDDC.
The first events that opened the way to SDDC were the concepts of software-defined networks (Software Defined Network, SDN), and why a little later software-defined storage systems (Software Defined Storage, SDS). In conjunction with server virtualization technology, which allowed emulating virtual machines based on physical servers with the necessary characteristics, network and storage virtualization made it possible to create a technology stack, now called SDDC.
This way of organizing the infrastructure spurred the development of cloud technologies that required flexibility and easy scalability. The cloud, as a model of providing various IT services and resources (XaaS) with remote access via the Internet, needed a technology to automate actions to deploy, reconfigure, and monitor computing resources without reference to static hardware. This opportunity was provided by software included in the technology stack for software-defined data centers.
Cloud4Y's corporate cloud provider uses the software-defined infrastructure of its cloud (Software Defined Infrastructure, SDI) to create a proprietary software package from VMware, so we will use it as an example. Despite the high cost of software, it provides the necessary for Enterprise-customers reliability and stability.
At the heart of the VMware vSphere platform, which, according to many experts, is the industry standard for enterprise infrastructures. vSphere includes an ESXi host and storage management server, the so-called vCenter Server. ESXi hosts are physical servers with VMware ESXi hypervisors that are responsible for emulating virtual machines, each on its own server. The vCenter Server "masters" ESXi-hosts for itself and is a "command post" that allows you not only to manage all the capabilities of each host in the group, but also to create clusters with options for balancing (DRS), live migration (vMotion) and fast recovery virtual machines on redundant hosts (HA) for fault tolerance.
Also, building a cluster will require a storage system connection. VMware solutions for software-defined storage systems ( Software Defined Storage, SDS ) data are also available - Virtual Volumes and Virtual SAN.
The cluster file system stores partitions with virtual machine files that can be read and written by all ESXi hosts in the cluster. Due to the storage in one place and the independence of the virtual machine from the physical platform, fast movement and recovery is achieved using HA, DRS, FT, vMotion .
For a minimal implementation of SDDC, it remains to mention the VMware solution for software- defined networks ( Software Defined Network or SDN ). VMware NSX replicates the network model at the software level. An NSX network is a library of logical network elements and services, such as logical switches, routers, firewalls, load balancers, VPNs, and others. Thanks to this solution, which is part of the vSphere platform, the administrators of the provider and the tenants of the cloud can connect virtual machines in the network with any topology as soon as possible.
Clients do not use VMware vCenter Server. The provider is responsible for managing the clusters and the physical equipment. Customers get a significant amount of control over their virtual data center using the convenient VMware vCloud Director self-service portal, about which we have already written . Creating an SDDC or vCOD for a client occurs in the shortest possible time from one vCloud Director panel, while creating the required number of virtual machines with the necessary characteristics and operating systems, and then they can be combined into internal networks or hybrid clouds.
vCloud Director and vCenter are part of the SDDC stack, which are responsible for automation and management, largely because of this, this solution is gaining popularity, saving IT-specialists from the monotonous routine work of administering the virtual machine park by automating many processes.
Also, many consider it necessary to allocate information security tools as a separate element of the stack, which should be virtualized and provided as services ( Software-defined security, SDsec ). In the future, we will devote a separate article to this, but for now let us summarize, giving a definition.
Using the VMware technology stack for implementing Cloud4Y's SDI cloud allows you to improve the quality of service (Quality of Service, QoS) and reduce the cost compared to implementing the required customer infrastructure options using traditional data centers by renting physical servers or buying your own. Profitability is achieved due to the refusal of customers from capital expenditures in favor of operating expenses and payment only for actually used resources with the possibility of hourly charging. Despite the commoditization of equipment to ensure the availability of 99.982% per month, fixed in SLA, Cloud4Y uses reliable equipment: server (HPE), network equipment (Cisco, Juniper), All-flash storage. The equipment is located in the network of TIER III data centers connected by an optical ring with duplication of communication channels.
All resources are dedicated. We set the minimum permissible performance of the CPU and RAM system. The number of MIPS per vCPU is at least 2900, which guarantees customers the claimed processor speed. Also, “re-signing” of physical RAM is not allowed, RAM Swaped is 0%. This means that the Configured Virtual RAM allocated during creation of the virtual machine, which the guest OS will see, is 100% allocated physical memory that is available to the virtual machine at any time. This approach in practice allows you to avoid performance degradation of cloud servers in cases where the operating system accesses blocks in RAM and expects a quick response, but these blocks are read and written to the hard disk, which is significantly inferior in speed, due to the load of other clients.
Free unlimited 5 Mbps unlimited channel with VPN and expandable up to 1000 Mbps is provided.
Technical support is provided in the mode of 24 * 7 * 365. The support is divided into three lines:
All three lines operate around the clock. At the core is quality management, which provides a steady way out of our business processes. This and other organizational and technical measures allow us to guarantee our clients a high level of security and quality of services. Having a stable and well-functioning system, we are ready to assume responsibility, including legal, and become a reliable link in the chain of processes of organizations of all levels and sizes.
To explore the cloud capabilities on the VMware stack, we offer free test access .
Hosting as a business is relatively simple. There are several options for the creation and development. You can become a reseller of a large hosting company or rent a dedicated server, or purchase your own equipment and place it in a data center. You will also need to select, install, and configure a web hosting control panel that allows you to manage software, such as a web server, database server, mail server, and other related programs through a web interface. It remains to solve the problem of technical support, start marketing and find customers. And it seems that nothing special, but you need to meet the competitive price, conditional "99 rubles per month", and therefore have to save.
')
What saves hosting discounters and business risks
Loukosterov use inexpensive and often problematic equipment. How to buy an expensive server and storage, but at the same time meet the competitive price? On Habrahabr, you can read about the case of failure by a major provider in the request to replace the old disk with a new one and install it under the guise of a new disk of the same old one. Disks “fall off” and you lose data. Everyone can independently assess the value of their corporate data, as a rule, it exceeds the amount of savings by at least an order of magnitude.
Partially solve the problem can backup. But this service is provided not by all cheap hosting, and if it offers to make copies of data, then perhaps it will be the same disk on which your working data is stored. In the event of a disk crash, your data and backup files are lost.
The second important component in the work of hosting is technical support. A 24-hour support service will require the hiring of qualified staff and a substantial payroll. What if the owner of the hosting decides that he will support the operation of the hosting on his own? 24 * 7 mode is canceled, because you also need to sleep. Problems happen at the wrong moment, customers are forced to wait and suffer losses.
The choice of a hosting provider in this market segment is complicated by the lack of a free trial period. The client is not able to first try the service, and then pay. Maybe someone will be lucky and there will be no problems with hosting, but first they have to take risks, since the low-cost airlines ask for 100% prepayment. In case of failure, money back programs will partially cover monetary losses, but not wasted time.
The business has needs that many hosting discounters cannot meet. When choosing a hosting you need to check
- does the hosting provide invoices for payment and closing documents for legal entities from Russia;
- is there a choice of pre-installed operating systems and is it possible to independently install your own OS;
- Are additional IP addresses provided?
- is there any protection against DDoS attacks on IT infrastructure, because an attack on neighbors can lead to total downtime.
So, cheap hosting is the lack of guarantees. This is how economic systems are structured, which is all due to something or someone else’s expense. If customers pay only "99 rubles per month", there is no possibility in the framework of commercial activities to provide services for a large amount. So you have to save on hardware, technical support and limited to a narrow functionality of the software.
There are projects for which this service will prove to be an ideal ratio of price and needs to be satisfied. However, if greater reliability, safety and flexibility is required, there are cloud products that, with a significantly higher price, have a higher customer value / price ratio. Comparing such services should, of course, be compared with similar services, and not with cheap hosting.
To better understand how the business of a cloud provider is structured, let's start by analyzing the concept of SDDC.
What is SDDC or FAQ about software-defined data centers
If we translate Software Defined into Russian, we will get the term "software-defined" or "software-configurable." However, the trend of shifting the focus from “hardware” to “software” better describes the first option, therefore we will use it as the main one.
So, Software Defined Data Center or SDDC is a software-defined data center. This concept was first proposed by Stephen Herrod in 2012, who was at the time the position of VMware Technical Director.
Such a late emergence of the SDDC concept is associated with the processes taking place in the IT industry. In the 1950s, the cost of hardware was disproportionately higher than the cost of programs, which were considered mere additions to hardware. That is, the first task was to develop a hardware solution, and a secondary one - software.
Gradually, the process of komoditization or the transition to the creation of more universal and interchangeable computers led to the understanding that the program can be used on any computer equipment. As a result, by the age of 80, software development becomes an independent industry, and the cost of software first approaches the cost of developing hardware, and then in some cases begins to exceed. This overrides the primacy of hardware in favor of software. All the specialized functionality is transferred to the software, and the hardware becomes more versatile. The process, extending to all elements of the data center system, creates prerequisites for the creation of software-defined data centers or SDDC.
The first events that opened the way to SDDC were the concepts of software-defined networks (Software Defined Network, SDN), and why a little later software-defined storage systems (Software Defined Storage, SDS). In conjunction with server virtualization technology, which allowed emulating virtual machines based on physical servers with the necessary characteristics, network and storage virtualization made it possible to create a technology stack, now called SDDC.
This way of organizing the infrastructure spurred the development of cloud technologies that required flexibility and easy scalability. The cloud, as a model of providing various IT services and resources (XaaS) with remote access via the Internet, needed a technology to automate actions to deploy, reconfigure, and monitor computing resources without reference to static hardware. This opportunity was provided by software included in the technology stack for software-defined data centers.
On the VMware stack and in secure data centers
Cloud4Y's corporate cloud provider uses the software-defined infrastructure of its cloud (Software Defined Infrastructure, SDI) to create a proprietary software package from VMware, so we will use it as an example. Despite the high cost of software, it provides the necessary for Enterprise-customers reliability and stability.
At the heart of the VMware vSphere platform, which, according to many experts, is the industry standard for enterprise infrastructures. vSphere includes an ESXi host and storage management server, the so-called vCenter Server. ESXi hosts are physical servers with VMware ESXi hypervisors that are responsible for emulating virtual machines, each on its own server. The vCenter Server "masters" ESXi-hosts for itself and is a "command post" that allows you not only to manage all the capabilities of each host in the group, but also to create clusters with options for balancing (DRS), live migration (vMotion) and fast recovery virtual machines on redundant hosts (HA) for fault tolerance.
Also, building a cluster will require a storage system connection. VMware solutions for software-defined storage systems ( Software Defined Storage, SDS ) data are also available - Virtual Volumes and Virtual SAN.
The cluster file system stores partitions with virtual machine files that can be read and written by all ESXi hosts in the cluster. Due to the storage in one place and the independence of the virtual machine from the physical platform, fast movement and recovery is achieved using HA, DRS, FT, vMotion .
For a minimal implementation of SDDC, it remains to mention the VMware solution for software- defined networks ( Software Defined Network or SDN ). VMware NSX replicates the network model at the software level. An NSX network is a library of logical network elements and services, such as logical switches, routers, firewalls, load balancers, VPNs, and others. Thanks to this solution, which is part of the vSphere platform, the administrators of the provider and the tenants of the cloud can connect virtual machines in the network with any topology as soon as possible.
Clients do not use VMware vCenter Server. The provider is responsible for managing the clusters and the physical equipment. Customers get a significant amount of control over their virtual data center using the convenient VMware vCloud Director self-service portal, about which we have already written . Creating an SDDC or vCOD for a client occurs in the shortest possible time from one vCloud Director panel, while creating the required number of virtual machines with the necessary characteristics and operating systems, and then they can be combined into internal networks or hybrid clouds.
vCloud Director and vCenter are part of the SDDC stack, which are responsible for automation and management, largely because of this, this solution is gaining popularity, saving IT-specialists from the monotonous routine work of administering the virtual machine park by automating many processes.
Also, many consider it necessary to allocate information security tools as a separate element of the stack, which should be virtualized and provided as services ( Software-defined security, SDsec ). In the future, we will devote a separate article to this, but for now let us summarize, giving a definition.
SDDC is an integrated level of abstraction that defines the entire data center using software tools, consisting of infrastructure components - processors, storage systems, network equipment and information security, automation and management tools that form common pools of virtual resources provided to consumers as services ( XaaS).
Using the VMware technology stack for implementing Cloud4Y's SDI cloud allows you to improve the quality of service (Quality of Service, QoS) and reduce the cost compared to implementing the required customer infrastructure options using traditional data centers by renting physical servers or buying your own. Profitability is achieved due to the refusal of customers from capital expenditures in favor of operating expenses and payment only for actually used resources with the possibility of hourly charging. Despite the commoditization of equipment to ensure the availability of 99.982% per month, fixed in SLA, Cloud4Y uses reliable equipment: server (HPE), network equipment (Cisco, Juniper), All-flash storage. The equipment is located in the network of TIER III data centers connected by an optical ring with duplication of communication channels.
All resources are dedicated. We set the minimum permissible performance of the CPU and RAM system. The number of MIPS per vCPU is at least 2900, which guarantees customers the claimed processor speed. Also, “re-signing” of physical RAM is not allowed, RAM Swaped is 0%. This means that the Configured Virtual RAM allocated during creation of the virtual machine, which the guest OS will see, is 100% allocated physical memory that is available to the virtual machine at any time. This approach in practice allows you to avoid performance degradation of cloud servers in cases where the operating system accesses blocks in RAM and expects a quick response, but these blocks are read and written to the hard disk, which is significantly inferior in speed, due to the load of other clients.
Free unlimited 5 Mbps unlimited channel with VPN and expandable up to 1000 Mbps is provided.
Technical support is provided in the mode of 24 * 7 * 365. The support is divided into three lines:
- The first line of technical support deals with issues of accessibility and technical issues up to the OS level inclusive, gets tickets and advises clients. In the case of queries that do not affect the data and do not relate to finance, customer issues can be resolved by any of the available methods.
- The second line of technical support deals with the level of hypervisors and operating systems, customization of client software, customization, log analysis, integration issues and deep support for customer information systems.
- The third line of support solves problems globally at the network level, at the level of data storage systems, and is responsible for the architecture and availability in general.
All three lines operate around the clock. At the core is quality management, which provides a steady way out of our business processes. This and other organizational and technical measures allow us to guarantee our clients a high level of security and quality of services. Having a stable and well-functioning system, we are ready to assume responsibility, including legal, and become a reliable link in the chain of processes of organizations of all levels and sizes.
To explore the cloud capabilities on the VMware stack, we offer free test access .
Source: https://habr.com/ru/post/343300/
All Articles