Bare metal servers presuppose strict performance control and ability to handle processor-demanding and disk I/O-intensive workloads. The customer chooses the model and configuration of a server, and we arrange deployment in any of our data centers within the next 40 minutes or less.
We prioritize performance and hardware quality over low prices. We aim to find smart solutions and automatize most routine procedures. Thus, we have come up with a ‘Post-cloud bare metal’ concept. We do not match cloud servers against bare metal ones. Our ‘Post-cloud bare metal’ approach combines high power and controllability of dedicated servers with cloud flexibility.
Fast automated provisioning
One of our smart solutions is automated provisioning. Deployment takes most hosting providers one or two days, or even a few weeks after the payment is processed if one makes a substantial order. Our automated provisioning reduces deployment time to 40 minutes. You do not have to worry about scalability of your projects, as we always keep hundreds of available servers that have been installed in our open racks and integrated into our intelligent hosting platform ready. Server and network performance will be automatically checked. We have never had a single case of providing our customers with non-functioning servers.
We are extremely attentive to detail. Thus, if you already have running servers and you order one more, our platform automatically selects one of the available servers that is the closest to the ones you already have. This speeds up support processes and reduces latency within your architecture.
We have simplified the process of ordering as well. No need to call, no need to bother with correspondence and paperwork. A few clicks, and you can count down to the launch of your new server. We do our best to manage our resources efficiently. That’s why not only have we automated provisioning, but deprovisioning process as well. Moreover, the servers that are currently not in use stay off. As there are no extra costs for electricity and amortization, it is possible to further reduce the price of bare metal.
All our bare metal servers are equipped with several network cards with the total bandwidth of 40 Gbps, with 20 Gbps uplink to the private network, and 20 Gbps uplink to a public network.
Server network interfaces are connected to two separate leaf-switches in different racks. Each leaf-switch is connected to two spine-switches, and each spine-switch is connected to two routers. Such a switching scheme provides 100% accessibility of the server.
Should a network card fail, it will never affect the network functions, as every network card is configured so that it can handle the entire volume of tasks. In case of failure, intelligent platform will detect a problem immediately, and support will fix it. You will know that the problem occurred only after it is resolved, and it will never affect the productivity of your infrastructure.
One of our goals is to provide our customers with simple and user-friendly tools, so that they could personalize their settings autonomously. Customers know their needs and demands best, that is why we have automated as many processes as possible and gave our clients an opportunity to customize plenty of settings. Cloud servers are not an exception; moreover, the concept itself presupposes flexibility that includes nearly immediate scalability, pay-as-you-go billing and fast deployment.
When we say, 'fast deployment', we mean it. Our intelligent platform monitors and updates if necessary server OS images on a daily basis. During deployment, the latest version of an OS with the latest security updates is installed on a server. The customer receives fully functioning servers that do not require installation of any updates that can sometimes be rather time-consuming.
Open source API
Our cloud operates on OpenStack with KVM hypervisor. OpenStack is a popular open platform with flexible architecture and an extensive ecosystem of ready-made integrations.
OpenStack API support in all our cloud products allows our clients to work with different modules of the platform with help of standardized user-friendly tools.
100% SSD storage
Everyone needs stable, productive and 'fast' clouds. A 'fast' cloud should have low DC-client and DC-DC latency, high server uplink, fast storage I/O, etc. Our cloud servers have local SSD-disks. SSD delivers fast I/O for demanding enterprise applications. With us, SSD is not a paid option but a standard for all cloud servers.
40 Gb network
While designing the concept of our cloud, we were fully aware that if we use SSD-disks in our servers, the other elements of the system should be adequate. The speed of the entire system is affected by the slowest element. Every cloud node has the total bandwidth of 40 Gb: 20 Gb to a private network, and 20 Gb to the public one. Our customers get a fast network and consistent cloud performance and they do not have to pay for any ‘extra’ features.
Free Global and Local Private Networks
You can combine your cloud servers, bare metal servers and cloud storages in an integral Free Global Private Network. Our global private network allows for effective resource management, including redistributing workloads and power, and setting data replication between data centers.The network's traffic is not rated. It is unlimited and free.
Cloud servers located in one area can be combined in a local private network. The data in a local private network is securely protected, and the traffic within that network is free as well. Private networks are isolated from public ones on a hardware level, that is why there is no risk of data breach or leakage.
When one works with data, they cannot fully rely even on the most reliable system. This is why we have implemented a snapshot cloud data backup system. A snapshot is a point-in-time copy of your data. With help of snapshots, one can back up big volumes of data, and this will not affect the productivity of an application. Once you set the snapshot protocol parameters, it will keep backing up your data automatically.
Urgently addressing consequences of an accident, saving customers’ data, figuring out the causes of data corruption or loss are not the things with which we are eager to start our day. Frankly, we would prefer never to have to deal with such issues at all.
We would like to make our cloud storage as reliable as possible and to reduce possible risks and accidents to a minimum. Thus, we have created a storage, where most processes are fully automated and protected from the influence of human factor, and where the critical processes are repeatedly reserved. We believe that we have succeeded in our intentions, as after the initial setup, the storage works without requiring customers’ attention.
Fast and fault-tolerant
We use OpenStack Swift to store your data. To reduce possibility of data loss, we use triple replication. Whatever happens at any time, three copies of your data are stored on separate storage nodes.
Response time is a crucial parameter for many services and applications, and it depends on many factors, including storage location. We thought about it in advance and made geo-replication settings very simple, so that our customers could deal with automatic geo-replication to another area themselves. They need literally just a few clicks. Furthermore, geo-replication between remote areas is free, as the traffic within our Free Global Private Network is not rated.
If there is data, there must be backups. Nobody expects data corruption or data loss, however restoring data is far more time-consuming and difficult than backing up. We offer the choice between Duplicity, Veeam and Cloudberry backup tools.
Three defense lines of your data
SSL. SSL protocol, which is one of the comprehensive security features that we use, is responsible for the security of your data. SSL-encryption increases the level of sensitive data protection and prevents potential data leakages.
Objects security. Along with secure storage of our customers’ data, we have thought of secure sharing. Sometimes data stored on a private account has to be accessed from elsewhere. For such purposes, we have a Temporary URL tool. A Temporary URL can be valid for one hour, one day, one week, or one month. The shorter the validity term is, the smaller are the risks of data leakage. Furthermore, the time of access to the stored data is strictly limited.
Data protection. As humans tend to make errors, it is possible to set object versioning and even trash versioning in our storage. In this case, a set amount of both, live- and trash-versions of your data will be stored. Accidental copying of a file to a wrong folder, as well as accidental deletion, will not lead to complete data loss. Your files can be always restored or backed up.
Our goal is to provide you with user-friendly and functional account administration and data management tools. of course, we have our own feature-rich API. Yet, if you have to work with legacy software, or whatever the reasons may be, we offer you FTP access, as well as CyberDuck and Web File Manager tools support.