Backups are different. Today I will talk about Acronis Backup Cloud (BaaS)
In this article, I will step by step describe the operation of the Acronis Backup Cloud backup service (formerly known as Acronis Backup as a Service), developed by Acronis engineers. I’ll tell you what “backups as a service” is inside, and how it all works. I turn directly to the description of the service itself.
The service consists of server and client parts. The “agent” and “agentless” technologies are used, depending on the infrastructure. An agent is installed on the client computer or host virtualization server, whose task is to connect the client computer / host to the Acronis Backup Cloud server and perform backup and recovery tasks.
Here are the Agents (Clients) developed:
Let's look at how the server side is arranged. The server part of the service consists of two components - system management and backup storage.
The control component is accessible via the Internet and allows using a web browser to manage the backup of remote machines and already created backups; Create, edit, and delete backup policies and retention policies. configure encryption of created backups using AES or GOST standard, and if necessary, save separate backups locally; monitor the status of remote machines; restore individual files / folders, disks / partitions, or the entire machine from the cloud directly to bare metal. One of the most distinctive and useful features is the creation of a hierarchy of subordinate administrator and user accounts, which distribute access to data and remote machines.
The storage component enables you to deploy scalable, low-cost yet reliable data storage. Backup storage consists of a group of servers on which client data is written. To ensure a sufficient level of reliability of all stored user data, each input file is divided into “K” blocks and then “NK” (where N is a number greater than K) redundancy blocks using the Reed-Solomon error correction algorithm are added. All blocks are stored independently of each other, and the safety of any “K” blocks from the recorded “N” guarantees the restoration of stored user data.
Acronis storage is built on physical servers. But at the same time server roles are assigned directly to disks.
There are three server roles: Metadata Server (MDS), Storage Server (STS) and Front-end Server (FES).
The Metadata Server is responsible for storing information about the fragments into which the file is divided, and the location of these fragments on the servers. This is the most critical component of the system.
To ensure high availability and fault tolerance of the storage, it is recommended to have several servers with the MDS role. One of the servers becomes the primary, and metadata is periodically replicated to other servers with the MDS role.
In addition, a System Management Component (MGMT) is also installed on each server that has the MDS role. If the primary MDS server stops working, the system management component is automatically enabled on another server with the MDS role, so the storage management web console is always available.
Storage Server (STS) is designed to store pieces of data.
The Front-End server allows Acronis Backup Cloud customers to access the storage and transfer data between the user side and the Acronis data warehouse.
Now let's see how it works in practice. Administrators of IT-Lite (the service provider) and end-user companies are given access to manage the service through the web console. The diagram below shows the standard backup service architecture. Blue arrows indicate interactions between software components. The black arrows show how administrators and end users gain access to the backup service.
IT-Lite administrators have access to manage groups and user accounts.
Administrators of end-user companies have rights that allow you to manage users who are only in their group! And end users, in turn, gain access to the console, in which you can add computers and create an automatic backup schedule. The service is integrated with the site of the service provider, which allows users to automatically register after filling out the form on this page.
For the end user, working with the system is simple and intuitive. After creating a user account and entering the Personal Account, you must specify the computer for which the backup task will be configured. To do this, click on “+” and select the OS on which the agent will be installed. The next step is to configure the task for backup.
The User Management Console displays all the Tasks that were previously performed for a specific computer; in addition, there is the “Create a New Task” and “Restore Data from the Cloud” functions. In addition, viewing the current backup status for a given computer is available.
BaaS service user data is stored on servers located in a certified TIER-3 class data center.
Thanks to the architecture used, it provides protection against failure at the level of individual servers and individual disks, which, we note, is impossible when using RAID arrays, which are used to create fault-tolerant storages. The system also uses a full data integrity check. The redundancy level is configured in the storage management console. The self-healing design used in the development of the storage allows avoiding losses in system performance typical of RAID arrays.
If one of the disks or even the whole server fails, the system will automatically rebalance, which avoids the immediate replacement of failed components.
I forgot to mention the peculiarity of using backups as a service - in case of an emergency it is necessary to restore the entire computer (virtual machine). Yes, indeed, if we compare the server recovery speed, then the local backup, of course, is faster. But in this case, a choice must be made - either in favor of speed, but without a guarantee from the jambs, or in favor of the reliability of the solution, but you will have to sacrifice a little speed.
To summarize this - I came to the conclusion that the question of the reliability / simplicity of local backups and the use of BaaS is rather a value issue. The one to whom the information is really valuable is more likely to choose a service, and the rest is the first option. But this is my personal opinion, it is not directly relevant to the matter.
Summarizing all of the above, I would like to add that, as they say, it is better to see once than hear a hundred times. Therefore, it is better to work with the service yourself and draw your own conclusions, since there is a free test .
How is the service arranged
The service consists of server and client parts. The “agent” and “agentless” technologies are used, depending on the infrastructure. An agent is installed on the client computer or host virtualization server, whose task is to connect the client computer / host to the Acronis Backup Cloud server and perform backup and recovery tasks.
About the client part
Here are the Agents (Clients) developed:
- Backup clients for Windows, Linux, and Mac - are responsible for backing up data on machines running Windows / Linux / Mac.
- Backup client for VMware - backs up VMware vSphere virtual machines (ESX / ESXi) without installing the corresponding clients on the guest systems. This client is installed on Windows machines with access to vCenter Server and storage, in which virtual machine backups are saved.
- Backup client for Hyper-V - backs up virtual machines on the MS Hyper-V platform without installing the corresponding clients on guest systems. This client installs on the Hyper-V host.
- Backup Client for MS Exchange Server - Microsoft Exchange Server 2003, 2007/10, 20013 versions are supported.
- Backup Client for MS SQL Server - Microsoft SQL Server 2005, 2008, 2008 R2, 2012, 2014 versions are supported.
About the server side
Let's look at how the server side is arranged. The server part of the service consists of two components - system management and backup storage.
The control component is accessible via the Internet and allows using a web browser to manage the backup of remote machines and already created backups; Create, edit, and delete backup policies and retention policies. configure encryption of created backups using AES or GOST standard, and if necessary, save separate backups locally; monitor the status of remote machines; restore individual files / folders, disks / partitions, or the entire machine from the cloud directly to bare metal. One of the most distinctive and useful features is the creation of a hierarchy of subordinate administrator and user accounts, which distribute access to data and remote machines.
The storage component enables you to deploy scalable, low-cost yet reliable data storage. Backup storage consists of a group of servers on which client data is written. To ensure a sufficient level of reliability of all stored user data, each input file is divided into “K” blocks and then “NK” (where N is a number greater than K) redundancy blocks using the Reed-Solomon error correction algorithm are added. All blocks are stored independently of each other, and the safety of any “K” blocks from the recorded “N” guarantees the restoration of stored user data.
Server roles
Acronis storage is built on physical servers. But at the same time server roles are assigned directly to disks.
There are three server roles: Metadata Server (MDS), Storage Server (STS) and Front-end Server (FES).
The Metadata Server is responsible for storing information about the fragments into which the file is divided, and the location of these fragments on the servers. This is the most critical component of the system.
To ensure high availability and fault tolerance of the storage, it is recommended to have several servers with the MDS role. One of the servers becomes the primary, and metadata is periodically replicated to other servers with the MDS role.
In addition, a System Management Component (MGMT) is also installed on each server that has the MDS role. If the primary MDS server stops working, the system management component is automatically enabled on another server with the MDS role, so the storage management web console is always available.
Storage Server (STS) is designed to store pieces of data.
The Front-End server allows Acronis Backup Cloud customers to access the storage and transfer data between the user side and the Acronis data warehouse.
Work with the service
Now let's see how it works in practice. Administrators of IT-Lite (the service provider) and end-user companies are given access to manage the service through the web console. The diagram below shows the standard backup service architecture. Blue arrows indicate interactions between software components. The black arrows show how administrators and end users gain access to the backup service.
Announce the rights, please!
IT-Lite administrators have access to manage groups and user accounts.
Administrators of end-user companies have rights that allow you to manage users who are only in their group! And end users, in turn, gain access to the console, in which you can add computers and create an automatic backup schedule. The service is integrated with the site of the service provider, which allows users to automatically register after filling out the form on this page.
Acronis Backup Cloud through the eyes of the user
For the end user, working with the system is simple and intuitive. After creating a user account and entering the Personal Account, you must specify the computer for which the backup task will be configured. To do this, click on “+” and select the OS on which the agent will be installed. The next step is to configure the task for backup.
The User Management Console displays all the Tasks that were previously performed for a specific computer; in addition, there is the “Create a New Task” and “Restore Data from the Cloud” functions. In addition, viewing the current backup status for a given computer is available.
Comparison with local backups
BaaS service user data is stored on servers located in a certified TIER-3 class data center.
Thanks to the architecture used, it provides protection against failure at the level of individual servers and individual disks, which, we note, is impossible when using RAID arrays, which are used to create fault-tolerant storages. The system also uses a full data integrity check. The redundancy level is configured in the storage management console. The self-healing design used in the development of the storage allows avoiding losses in system performance typical of RAID arrays.
If one of the disks or even the whole server fails, the system will automatically rebalance, which avoids the immediate replacement of failed components.
Backup locally - yes, faster, but less reliable
I forgot to mention the peculiarity of using backups as a service - in case of an emergency it is necessary to restore the entire computer (virtual machine). Yes, indeed, if we compare the server recovery speed, then the local backup, of course, is faster. But in this case, a choice must be made - either in favor of speed, but without a guarantee from the jambs, or in favor of the reliability of the solution, but you will have to sacrifice a little speed.
To summarize this - I came to the conclusion that the question of the reliability / simplicity of local backups and the use of BaaS is rather a value issue. The one to whom the information is really valuable is more likely to choose a service, and the rest is the first option. But this is my personal opinion, it is not directly relevant to the matter.
Summary
Summarizing all of the above, I would like to add that, as they say, it is better to see once than hear a hundred times. Therefore, it is better to work with the service yourself and draw your own conclusions, since there is a free test .
