“Closer to the people”: how to make IaaS more accessible
According to data from one of Cisco ’s whitepaper , by the end of this year, more than 70% of data processing tasks will be solved by IaaS providers . Today we will talk about what approaches allowed us to make the IaaS friendly and convenient.
/ Flickr / Dennis Skley / CC
Studies of various analytical companies and experts only support the growth forecasts of the data center. For example, Statista talk about increasing the annual traffic of data centers to 9.8 exabytes in 2018. Such loads exacerbate the tasks of efficient use of resources, increasing the flexibility of the IT infrastructure and timely response to customer requests.
')
The solution can be found much faster and at lower cost if the company is ready to use the tools that it offers to its customers. We apply this approach in the management of our own IaaS-provider - almost all of our employees work with the available service tools and know in practice what and how it works.
Feedback from the team allows you to sensibly assess the situation and the need for those or other opportunities that are necessary for the effective work of clients and the company as a whole. One example is the development and implementation of a load monitoring system for the infrastructure of our clients. With the help of such a service, a customer can estimate in real time what he pays for and refuse a part of unused resources.
To realize this opportunity, we didn’t need any complicated manipulations. What we did: we chose a graphical library for displaying information and configured vSphere integration using the .NET SDK to obtain load data for individual virtual machines. All this is one more step towards the pay-as-you-go model - paying only for consumed resources and entering the automatic scaling system at peak loads.
Analysis of the market and customer opinions led us to understand the need for input and other innovations. We have already talked about this more than once, but let us briefly dwell on the main points:
Such projects do not require a global revision of the entire service, multi-year design and development. At the same time, they help to increase the level of accessibility of our services to an audience that is sufficiently far from IT specificity and does not have specialized knowledge.
/ Flickr / Bruno Cordioli / CC
The physical condition of the infrastructure requires similar attention. Here you need to understand not only the immediate tasks, but also the general trends of IT development, which will shape the needs of customers.
Next, you need to think about the balance of performance, demand for resources and supporting ecosystems. All this is impossible without a rethinking of the approaches to organizing the work of the data center, optimizing risks and increasing the level of service availability.
This will not prevent the policy of openness . Following this principle, we have prepared and published relevant materials about the data center in which our equipment is located: SDN (St. Petersburg), Dataspace (Moscow) and AHOST (Alma-Ata). In addition, we have provided basic performance data .
It can be said that this principle is generally accepted for almost every aspect of our business. If we talk about the availability of our infrastructure, here we provide real financial guarantees and openly share our SLA . Plus, we talk about our troubleshooting experience and analyze incidents that occurred with Western colleagues.
In addition, we conduct photo tours of our “cloud” right here in the blog on Habré so that you can get acquainted with the real equipment as closely as possible. And even give recommendations on the choice of IaaS-provider so that you can pay attention not only to the level of availability, but also other nuances of cloud services.
/ Flickr / Dennis Skley / CC
Studies of various analytical companies and experts only support the growth forecasts of the data center. For example, Statista talk about increasing the annual traffic of data centers to 9.8 exabytes in 2018. Such loads exacerbate the tasks of efficient use of resources, increasing the flexibility of the IT infrastructure and timely response to customer requests.
')
The solution can be found much faster and at lower cost if the company is ready to use the tools that it offers to its customers. We apply this approach in the management of our own IaaS-provider - almost all of our employees work with the available service tools and know in practice what and how it works.
Feedback from the team allows you to sensibly assess the situation and the need for those or other opportunities that are necessary for the effective work of clients and the company as a whole. One example is the development and implementation of a load monitoring system for the infrastructure of our clients. With the help of such a service, a customer can estimate in real time what he pays for and refuse a part of unused resources.
To realize this opportunity, we didn’t need any complicated manipulations. What we did: we chose a graphical library for displaying information and configured vSphere integration using the .NET SDK to obtain load data for individual virtual machines. All this is one more step towards the pay-as-you-go model - paying only for consumed resources and entering the automatic scaling system at peak loads.
Analysis of the market and customer opinions led us to understand the need for input and other innovations. We have already talked about this more than once, but let us briefly dwell on the main points:
- redesign of the main sections of the company's website (and one more improvement in the future)
- development of a special calculator that allows you to quickly assess the potential costs of maintaining the infrastructure and move on to its launch
- setting up pre-installed software (control panel and the ability to track license status) and content management tools (pre-installed CMS)
- automation of various service tasks ( an example of our DNS manager)
Such projects do not require a global revision of the entire service, multi-year design and development. At the same time, they help to increase the level of accessibility of our services to an audience that is sufficiently far from IT specificity and does not have specialized knowledge.
/ Flickr / Bruno Cordioli / CC
The physical condition of the infrastructure requires similar attention. Here you need to understand not only the immediate tasks, but also the general trends of IT development, which will shape the needs of customers.
Next, you need to think about the balance of performance, demand for resources and supporting ecosystems. All this is impossible without a rethinking of the approaches to organizing the work of the data center, optimizing risks and increasing the level of service availability.
This will not prevent the policy of openness . Following this principle, we have prepared and published relevant materials about the data center in which our equipment is located: SDN (St. Petersburg), Dataspace (Moscow) and AHOST (Alma-Ata). In addition, we have provided basic performance data .
It can be said that this principle is generally accepted for almost every aspect of our business. If we talk about the availability of our infrastructure, here we provide real financial guarantees and openly share our SLA . Plus, we talk about our troubleshooting experience and analyze incidents that occurred with Western colleagues.
In addition, we conduct photo tours of our “cloud” right here in the blog on Habré so that you can get acquainted with the real equipment as closely as possible. And even give recommendations on the choice of IaaS-provider so that you can pay attention not only to the level of availability, but also other nuances of cloud services.
More of our materials on the topic:
- How to create a virtual infrastructure provider
- How to choose a direction for the development of an IT project
- Cloud Technology Myths: Part 1 , Part 2, and Part 3
- Personal experience: How we automated work with DNS records in the hosting panel
- A selection of materials about clouds, data centers and the development of services
- Acquaintance with IaaS-provider: what we write on Habré and in our blog
- Digest: IaaS provider operation, SSL certificates, data center and our “Friday format”
- "The little things that we noticed": case IaaS-provider
- “Now Required”: Issuing SSL Certificates with DNS Records
- “How it works”: introduction to SSL / TLS
Source: https://habr.com/ru/post/330714/
All Articles