Data Backup

Started by robicse, Oct 04, 2022, 07:53 AM

Previous topic - Next topic

robicseTopic starter

Data protection involves the presence of backup — backups from which you can restore them. For most companies and organizations, data backup is among the most important priorities. About half of the companies work with their data as a strategic asset.
And the value of stored data is constantly growing. They are used to improve the quality of customer service, support current activities, research and development, accounting, they are involved in automation systems, the Internet of Things, artificial intelligence, etc. Therefore, the task of protecting data from hardware failures, human errors, viruses and cyber attacks becomes extremely urgent.

There is an increase in cybercrime in the world. Last year, more than 75% of companies were subjected to cyber attacks. Compromising clients' personal data and confidential files can have serious consequences and lead to huge losses.
At the same time, there is a culture of working with data, an understanding that data is a valuable resource with which a company can make additional profits or reduce costs, and at the same time, a desire to ensure reliable protection of its data.

Store and protect

According to the survey results, about a quarter of respondents perform data backup on a monthly basis, the same number – weekly, and more than a quarter – daily. And this is quite justified: as a result of such foresight, almost 75% of organizations avoided downtime due to data loss last year. Improving software tools and services help them in this.

According to IDC's research of the global Data Replication and Protection software market, its sales in the world will grow by 4.8% annually from 2018 to 2022 and reach $8.9 billion.
Analysts in their report (Global Data Backup Software Market Growth 2019-2024), they concluded that in the next five years, the average annual growth rate of the global data backup software market will be 7.6%, and in 2024 its volume will reach $2.756 billion against $1.836 billion in 2021.

At the same time, the popularity of cloud backup is growing: sales of such products and services are projected to grow more than twice as fast as the data protection software market as a whole. According to Gartner's forecast, already this year up to 20% of enterprises will use cloud backup.

According to IKS Consulting, the segment "cloud backup as a service" (BaaS) is increasing by an average of 20% per year. According to the Acronis 2019 survey, companies are increasingly relying on cloud backup: more than 48% of respondents use it, and about 27% prefer to combine cloud and local backup.

Requirements for backup systems

Meanwhile, the requirements for data backup and recovery software are changing. In order to successfully solve data protection problems and optimize costs, companies are ready to purchase simpler, more flexible and inexpensive solutions, according to Gartner analysts. The usual methods of data protection do not always meet the new requirements.

Data backup and recovery systems should provide for simple deployment and administration, convenient management of the backup and recovery process, and prompt data recovery. Modern solutions often implement data replication functions, allow you to automate operations, provide integration with clouds, built-in archiving functions, and support hardware snapshots of data.

According to Gartner's forecast, in the next two years, up to 40% of companies will switch to new backup solutions, replacing existing software, and many will simultaneously use several products or services that optimally protect certain systems. Why are they not satisfied with the previous backup and data recovery solutions?

All in one

Analysts believe that as a result of this transition, companies get more flexible, scalable, simple and productive systems, often representing unified software for data management and storage. Advanced backup and recovery products include tools for effective data management, make it possible to move data to where its storage is most efficient (including automatically), manage, protect and restore it.

With the growing diversity and volume of data, comprehensive data protection and management becomes an important requirement: files, databases, data from virtual and cloud environments, applications, as well as access to various types of data in primary, secondary and cloud storage.

Comprehensive data management solutions provide unified data management across the entire IT infrastructure: their backup, recovery, archiving and snapshot management. However, administrators should clearly understand where, for how long and what data is stored, what policies apply to them. Fast recovery of applications, virtual machines and workloads from local or cloud data storage minimizes downtime, and automation minimizes errors due to the human factor.

Large organizations with a combination of legacy, traditional and modern applications often choose backup systems that support a wide range of operating systems, applications, hypervisors and relational databases, with high scalability (up to several petabytes and thousands of clients), as well as providing integration with a wide range of storage systems, public, private and hybrid clouds and tape drives.

As a rule, these are platforms with a traditional three-tier architecture of agents, media web servers and a management server. They can combine the functions of backup and recovery, archiving, disaster recovery (DR) and cloud backup, optimize performance using artificial intelligence and machine learning algorithms.

According to Forrester, centralized management of data sources, policies, the ability to reliably restore data and security are the most important characteristics of backup solutions.

Modern solutions can perform backups of virtual machines based on snapshots at any frequency with virtually no reduction in the performance of working environments. They bridge the gap between the Recovery Point Objective (RPO) and the Recovery Time Objective (RTO), guarantee the availability of data at any time and ensure business continuity.

Data volume growth

Meanwhile, the exponential growth of the volume of data generated continues in the world, and this trend will continue in the coming years. According to IDC's forecast, the volume of data created per year will grow from 2018 to 2025 from 33 to 175 ZB. The average annual growth rate will exceed 27%. This growth is also influenced by the increase in the number of Internet users. Last year, 53% of the world's population used the Internet. The number of Internet users increases by 15-20% annually.
 New and developing technologies, such as 5G, UHD video, analytics, It, artificial intelligence, AR/VR, entail the generation of increasingly large amounts of data. Sources of data growth are also entertainment content and video from CCTV cameras. For instance, the market for storing video from surveillance cameras, according to forecasts of MarketsandMarkets, will grow by 22.4% per year and will reach $19.28 billion this year.

and over the past two or three years, the volume of corporate data has grown by about an order of magnitude. Accordingly, the backup task has become more complicated. Data storage capacities reach hundreds of terabytes and continue to increase as data accumulates. The loss of even a part of this data can affect not only business processes, but also affect brand reputation or customer loyalty. Therefore, the creation and storage of backups significantly affects the entire business.

It can be difficult to navigate the offers of vendors offering their backup options. There are different options for creating and storing backups, but the most popular are local backup systems and the use of cloud services. Backup to the cloud or to the provider's data center provides reliable data protection and minimizes the risks associated with software failures, technical equipment malfunctions and employee errors.

Migration to the clouds

Data can be accumulated and stored in your own data centers, but at the same time you will have to ensure fault tolerance, clustering and capacity scaling, and have qualified specialists in the administration of storage systems.
In these conditions, outsourcing all such issues to a provider is very relevant. For instance, when web hosting databases in a provider's data center or in the cloud, you can assign responsibility for storing, reserving data, and operating databases to professionals. The provider will be financially responsible under the service level agreement. Among other things, this allows you to quickly deploy a typical configuration to solve a specific task, as well as provide a high degree of availability by reserving computing resources and backup.

At the same time, more and more data will be stored not in corporate networks and not on end devices, but in the cloud, and, according to IDC, the share of data in public clouds will grow to 42% by 2025. Moreover, organizations are moving to the use of multi-cloud infrastructures and hybrid clouds. This approach is already followed by 90% of European companies.

Cloud backup is a data backup strategy that involves sending a copy of the data over the network to a web server outside of its own web site. This is usually a service provider's server that charges the client based on the allocated capacity, bandwidth, or number of users.

Widespread adoption of cloud technologies and the need to manage large amounts of data contribute to the growing popularity of cloud backup solutions. In addition, the introduction of cloud backup solutions is associated with such advantages as simple management and monitoring, real-time backup and recovery, easy integration of cloud backup with other corporate applications, data deduplication and support for various clients.

Analysts consider Acronis, Asigra, Barracuda Networks, Carbonite, Code42 Software, Datto, Druva Software, Efolder, IBM, Iron Mountain and Microsoft to be the key players in this market.

Multi-cloud environments

Storage system vendors are doing everything possible to ensure that their products work effectively in a multi-cloud environment. The task is to simplify the use of data and move them to where they are needed, and their storage is the most efficient. For instance, they use next-generation distributed file systems that support a single namespace, providing access to data in different cloud environments, and offer common management strategies and policies in different clouds and at the local level. The ultimate goal is to manage, protect and efficiently use data wherever it is located.

Monitoring is another of the problems of multi—cloud storage. Monitoring tools are needed to track results in a multi-cloud environment. An independent monitoring tool developed for multiple clouds will allow you to get the big picture.

Combining peripheral and multi–cloud storage is also not an easy task. In order for these systems to work effectively together, it is necessary to know the volumes and types of data, where and how this data will be collected, transmitted and stored. To plan the process, you will also need to know how long each type of data should be stored, where, when and how much data will need to be transferred between different systems and cloud platforms, how they are backed up and protected.

All this will help administrators minimize the difficulties associated with combining peripheral and multi-cloud storage.

Data on the periphery

Another trend is peripheral computing. According to Gartner analysts, in the coming years, about half of all corporate data will be processed outside of traditional data centers or cloud environments: an increasingly significant proportion of them is located on the periphery — for storage and local analytics. According to IDC's forecast, the share of "peripheral" data in the EMEA region will almost double — from 11% to 21% of the total volume. The reasons are the spread of the Internet of Things, the transfer of analytics and data processing closer to their source.

Peripheral infrastructure – data centers of various scales and form factors - offer enough opportunities for data processing and storage and provide low latency. In this regard, changes are planned in the proportion of data volumes placed in the core of the network/data center, on its peripherals and on end devices.

The transition from cloud and centralized computing to peripheral computing has already begun. Such systems are becoming more and more in demand. The costs and complexity of creating a centralized architecture for processing large amounts of data are excessively high, such a system can become poorly managed compared to the distribution of data processing on the periphery or at the appropriate network level. In addition, data can be combined or depersonalized on the periphery before being sent to the cloud.

Data abroad

Some companies prefer to store data abroad, considering this option to be reliable protection of data from unauthorized access and an important risk reduction factor. Data abroad is a guarantee of the protection of valuable information.
The equipment located abroad is not under jurisdiction. And thanks to encryption, data center employees may not have access to your data at all. In modern foreign data centers, highly reliable equipment is used, high reliability indicators are provided at the level of the data center as a whole.

The use of foreign data centers can have a number of other advantages. The Client is insured against the risks associated with force majeure or unfair competition. The use of such platforms for data storage and processing will minimize such risks. For instance, in case of withdrawal of web servers, the company will be able to keep a copy of its systems and data in foreign data centers.

As a rule, the IT infrastructure of foreign data centers is quality standards, a high level of security and control of data storage. They use the latest IT solutions, firewalls, encryption technologies for communication channels, and means of protection against DDoS attacks. The power supply of the data center is also implemented with a high level of reliability (up to TIER III and IV).

web sites are often used as a backup data center. Thus, maximum safety and reliability are achieved, risks are minimized. In some cases, they are convenient for web  hosting data and connecting European customers to them. At the same time, the best response time for European users is achieved. Such data centers have direct access to European traffic exchange points.
 For instance, we offer our customers 4 data placement points in Europe at once — Zurich (Switzerland), Frankfurt (Germany), London (UK) and Amsterdam (Netherlands).

What should be considered when choosing a data center?

Using the services of commercial data centers, in addition to a convenient cost structure, the business receives a more flexible service that can be scaled in real time, and only consumed resources are paid (pay-per-use). The services of an external data center also allow you to reduce the risks associated with the uncertainty of the future, easily adapt IT to new technological trends, focus on your key business processes, and not on maintaining the IT infrastructure.

During the construction and operation of their web sites, providers take into account the best practices and international standards that place high demands on data center engineering and IT systems, such as ISO 27001:2016 Information Security Management (information security management), ISO 50001:2011 Energy Management System (effective planning of data center power supply systems), ISO 22301:2017 Business Continuity Management System (ensuring the continuity of data center business processes), as well as European standards EN 50600-x, PCI DSS standard, concerning the security of processing and storing data of plastic cards of international payment systems.

As a result, the customer receives a fault-tolerant service that provides reliable reliable data storage and continuity of business processes.


At the last job, backups were hourly, checking one of the daily allowances — once a day.
I understand this is not the answer that was expected, but I also wonder how many hourly backups?
Please clarify what is the gap between RTO and RPO, if these are actually orthogonal axes?


Deleting backup copies under the condition "older than N days" is a very bad idea. And if your working server breaks down for N+1 days? Yes, the first backup task after restoring the server will kill all your backups!!! And it's not a fact that by this time your work base will have time to recover.
It is much better to save backups by number, for example, we save "the last 10 backups". If there are more than 10 backups, then we delete the oldest one. And it is also good to link the deletion of backups to the successful completion of the backup, i.e. if the backup did not take place, we do not delete anything from the backups, we delete only if a fresh backup took place.

the backup server + local backups to a separate array (or even just hdd), hidden from the system using the built-in Windows backup system, proved to be a good approach. As far as I know, there are no known cases yet when cryptographers would get to such hidden volumes. But even if this happens, there is a copy on the backup web server.