Will robots take over data management? In recent years, backup and disaster recovery system vendors have introduced several significant innovations. But the best is yet to come. 

Modern data protection solutions, encompassing backup, disaster recovery, replication, and deduplication, are constantly evolving. Manufacturers have moved from a stage of manual configuration to automation. However, this is not the end of the road. There is increasing talk about the era of autonomous backup and even autonomous data management. Is this a near future reality, or just a fantasy?

Opinions on this matter are divided. Skeptics cite the example of autonomous cars. Although prototypes have appeared on the streets of San Francisco, the road to their widespread adoption seems to be a long way off. On the other hand, proponents point to robotic vacuum cleaners that are displacing traditional vacuum cleaners from homes. If humans can be eliminated from processes that require high precision, why not do the same in areas closely related to IT?

Automation and autonomy are very similar concepts, sometimes incorrectly used interchangeably. Nevertheless, there are some subtle differences between them. Automation means that the tasks performed are based on pre-defined parameters that must be updated as the situation changes. This is how elevators, office software, washing machines, robotic assembly lines, and most backup and DR systems work.

On the other hand, autonomous processes differ from automated ones in that they are constantly learning and adapting to the environment. In such cases, human intervention is not needed or is minimal. A great example is the aforementioned robotic vacuum cleaners or driverless cars.

The authors of the concept of autonomous data management assume that processes should take place invisibly, although under human control. Autonomy somehow combines automation with artificial intelligence (AI) and machine learning (ML), so that the data protection system intuitively adapts to the situation.

AI and ML technologies enable the automation of data management processes and minimize human intervention and supervision. Proponents of such a solution argue that it increases operational efficiency, extends uptime, improves security, and the level of services offered.

Clouds Force Change

If companies only stored data in on-premises environments, it would be possible to do without autonomous tools, but in the last two years, things have become much more complicated. Enterprises have moved some of their assets to the public cloud, which has contributed to the growing importance of hybrid and multi-cloud environments. It was supposed to be easier and cheaper, but the ongoing adoption of cloud services is causing sleepless nights for many IT managers.

The main problem lies in the excessive dispersion of data, which is located both in the local data center and in external service providers such as Amazon, Google, Microsoft, or smaller local providers. Managing, and especially protecting, digital assets scattered across various locations is a challenge. The situation is worsened by the relatively narrow range of vendors’ tools optimized for managing corporate data for hybrid and multi-cloud environments.

Part of the products provide support for multiple clouds through centralized control, although they consume many expensive resources. There are also efficient solutions, but only within a single cloud environment. Their main drawback is scalability in the clouds of different providers. In any case, in both of the aforementioned cases, operating costs are higher than desired.

Another problem is the excessive haste in implementing cloud technologies, leading to an increase in the number of point solutions. Cloud environment architects, application developers, and analysts implement independent data management solutions, which deepens the chaos and limits the possibilities of central management.

The data protection strategy in the cloud environment also leaves much to be desired. Security specialists emphasize that in today’s world, the most effective way to stop attackers is through preventive measures. Unfortunately, most modern technologies take a passive approach to resources stored in the cloud. In practice, this means that they create backups and restore backups after an attack, which results in unplanned downtime.

In summary, autonomous backup supports operations in multiple clouds, eliminates functional silos, automates all processes with minimal human intervention, and increases cyber resilience through active methods of detecting and preventing ransomware attacks.

To Limit the Role of the Weakest Link

It has long been known that people are the weakest link in the data protection system. This is particularly evident in environments that require fast and data-driven decision-making. It is also undeniable that people are prone to errors and slower than AI-based solutions, especially when it comes to mundane, repetitive tasks.

So will robots send IT department employees to the pasture in the near future? So far, no one is talking about it loudly. According to the authors of the concept of autonomous data management, the best solution in a complex, hybrid and multi-cloud environment is autonomous work. This means that data will self-optimize and repair itself, as well as move between different environments. Self-optimization uses artificial intelligence and machine learning to adapt to the principles and services related to data protection and management. Self-healing is the ability to predict, identify, and correct service errors or performance issues.

On the other hand, self-service assigns appropriate protection policies and manages and deploys applications and services without human intervention. What does this mean?

In the traditional model, a programmer deploying a new application relies on manual processes, which lengthens it. Autonomous data management eliminates all manual tasks, while protecting the application throughout the process, without the need for additional actions on the part of the application developer or IT staff.

Autonomous Data Management – Is It Worth It?

The concept of autonomous data management looks very promising. Importantly, some backup and DR system vendors are announcing the launch of such solutions in the near future, not in the coming years. On the market, you can already find products that use Machine Learning to early detect anomalies that signal an attempt to attack the backup system. Some companies also use partially AI-based solutions combined with DLP systems, which helps classify and tag information, and thus copy and protect the most important data.

However, only the widespread adoption of systems that provide autonomous data management will allow us to answer the fundamental question – is it worth the effort?

Some data protection specialists warn against excessive optimism. In their opinion, the biggest obstacle to the adaptation of autonomy in backup and DR processes may be the collection of a sufficiently wide range of data to be able to analyze various scenarios. It is difficult to imagine that vendors of solutions would share such information with each other.

It is also difficult to count on the openness of IT department employees, as they may fear that new products will deprive them of their jobs. It can also be safely assumed that the term “autonomy” will be overused by marketers, which on the one hand encourages customer investment, and on the other hand, threatens that low ratings of disappointed users will deter potential customers. It is possible that there will be limitations related to computing power, as well as the costs of such a solution. Nevertheless, it is worth closely following such initiatives, especially as it concerns large companies and institutions storing data in different environments.

Storware develops towards autonomous

While full autonomy might still be a distant goal, Storware’s focus on AI and automation is a significant step in that direction. These features have the potential to significantly improve efficiency, reduce human error, and enhance overall data protection.

In the near future, Storware will implement a number of improvements that will allow for:

• Automation: The Backup Assistant and conversational layer aim to automate routine tasks and provide intelligent responses, reducing human intervention.
• Intelligence: Storebrain’s ability to learn from collective data and provide optimal configurations demonstrates a move towards intelligent decision-making.
• Proactive Protection: The integration of AI into Isolayer for threat prevention showcases a proactive approach to data management, essential for autonomous systems.

However, key to achieving full autonomy would be further development in areas like:

• Self-healing capabilities: The system should be able to identify and resolve issues independently.
• Predictive analytics: Accurate forecasting of system behavior and potential problems.
• Continuous learning: The system should constantly improve its performance based on new data and insights.

Snapshots and backups are both crucial for data protection. However, to maximize their benefits, it’s essential to understand their capabilities.

As data volumes and value continue to grow, data has become an invaluable asset for businesses, governments, consumers, and cyber-criminals alike. Cyber-criminals will stop at nothing to steal information or block legitimate users from accessing it. Fortunately, organizations have various tools and methods to protect their data, including backups and snapshots. While these methods share some similarities, they are often mistakenly seen as interchangeable. This article will delve into the fundamental differences between backups and snapshots and how they can complement each other.

The Indispensability of Backups

Until recently, it was common to say that people were either backing up their data or were planning to do so. However, this saying is no longer accurate. It’s increasingly difficult to find individuals or businesses that don’t perform backups. Backups are typically created on a regular schedule (e.g., nightly or multiple times a day) and can include all files on a server, emails, or databases. By archiving data in backups, users are protected against accidental data loss caused by errors, accidental deletions, or other failures. This is why backups are often referred to as “security copies.”

There are several types of backups. The simplest is a full backup, which creates a complete copy of the data to a destination storage device. Other methods include differential and incremental backups. A differential backup only backs up data that has been added or changed since the last full backup. An incremental backup, on the other hand, uses the previous backup as a reference point rather than the initial full backup.

A full backup is a complete copy of the data. If each backup is 10TB, for example, it will consume an additional 10TB of storage. Creating a backup every hour would consume 100TB of storage in just 10 hours. For this reason, storing multiple versions of backups is not a common practice.

The Role of RPO

A challenge with backups is achieving a suitable Recovery Point Objective (RPO), which defines the maximum amount of data loss that can be tolerated and the maximum acceptable time between a failure and the restoration of a system to normal operation. Businesses have varying requirements—some may be satisfied with a 24-hour RPO, while others strive for an RPO as close to zero as possible. For example, losing even a small amount of data in manufacturing companies can lead to production line downtime, lost product batches, and significant financial losses.

Some businesses determine their RPO based on the cost of storage compared to the cost of data recovery. These calculations help determine the frequency of backups. Another approach is to assess risk levels. In this case, a company evaluates which data can be lost without significantly impacting the quality and continuity of its business.

Backups are not optimal for creating short recovery points. Snapshots are much better suited for this purpose, which is why the two technologies should be used together. Snapshots are the preferred solution when high RPO requirements must be met, such as in 24/7 environments like internet service providers.

Snapshots for Specialized Tasks

A snapshot is a point-in-time capture of stored data. Its main advantage is its creation time, which is typically measured in minutes or even seconds. Snapshots are usually created every 30 or 60 minutes and have minimal impact on production processes. They allow for quick recovery to previous file versions at multiple points in time. For example, if a system is infected with a virus, files, folders, or entire volumes can be restored to a state before the attack.

However, snapshots are often a feature of NAS or SAN storage and are stored on that storage. This means they occupy relatively expensive storage capacity, and if the storage fails, users lose access to recent snapshot copies. While individual snapshots do not consume much space, their combined size can increase, leading to additional processing costs during recovery. Therefore, it’s good practice to limit the number of stored copies. Experts recommend not storing snapshots for longer than the last full backup.

Furthermore, migrating a snapshot from one physical location to another does not allow for environment restoration, which is possible with backups. Since a snapshot is not a complete copy of the data, it should not be considered the sole backup and should be combined with backups. In summary, backups provide the ability to restore data over long RPOs, often quickly and in detail, down to the file level.

Types of Snapshots

While snapshot creation processes vary by vendor, there are several common techniques and integration methods.

• Copy-on-write: This method copies any blocks before they are overwritten with new information.
• Redirect-on-write: Similar to copy-on-write, but it eliminates the need for a double write operation.
• Continuous Data Protection (CDP): CDP snapshots are created in real-time, capturing every change.
• Clone/mirror: This is an identical copy of an entire volume.

Summary

Snapshots and backups have their strengths and weaknesses. Generally, backups are recommended for long-term protection, while snapshots are intended for short-term use and storage. Snapshots are typically useful for restoring the latest version of a server within the same infrastructure.

Both snapshots and file backups can be used together to achieve different levels of data protection, and this is actually the most recommended configuration for backup strategies.

Storware 7.0 – what’s new?

Storware 7.0 high level architecture:

Backup → Recover → Thrive

Data security and recovery are critical when dealing with virtualization tools. In the event of an outage or disaster, businesses must be able to rapidly restore virtual machines (VMs) and critical data. Being a virtualization management tool, oVirt not only shines in virtual environment administration but also provides virtual machine full backup and recovery capability.

This article will explore what oVirt is and how its backup and recovery system works.

What is oVirt?

Red Hat created oVirt, a robust open-source virtualization tool that uses the Kernel-based Virtual Machine (KVM) hypervisor.

With a web-based interface, this platform provides a centralized administration solution allowing consumers to manage compute, storage, and networking resources. oVirt facilitates companies’ design, management, and implementation of virtual machines.

oVirt is suitable for small and large companies since it is flexible and scalable. One important characteristic of oVirt is its ability to work with other open-source community projects such as Ansible for automation, Gluster for storage management, and PatternFly for user interface design. This integration helps customers to use existing tools in the opensource community and at the same time take advantage of oVirt’s advanced capabilities.

Components of oVirt

The oVirt Engine, hosts (nodes), and storage nodes make up the fundamental architecture of oVirt. These components form a comprehensive solution for managing virtualized environments.

• oVirt Engine

oVirt engine is a Wildfly-based Java application that functions as a web service. The engine talks to VDSM (Virtual Desktop and Server Manager) to deploy, start, migrate, and monitor VMs.

• Nodes

oVirt Nodes is a streamlined operating system built on CentOS. It operates on RHEL, CentOS, and Scientific Linux, utilizing the KVM hypervisor and the VDSM service, coded in Python. These nodes are Linux-based distributions with VDSM and libvirt installed. They also feature additional packages that simplify the virtualization of networking and system services.

• Storage Nodes

Storage nodes use either block or file storage which can be locally or remotely accessed through NFS (Network File System). These nodes are arranged into storage pools, offering options for high availability and redundancy.

The Latest oVirt Release Features

oVirt announced the release of a new update which became available on December 1, 2023. The new release named oVirt 4.5.5 is available on x86_64 architecture for:

• oVirt Node NG (based on CentOS Stream 8)
• oVirt Node NG (based on CentOS Stream 9)
• CentOS Stream 8
• CentOS Stream 9
• RHEL 8 and derivatives
• RHEL 9 and derivatives
• Experimental builds are available for ppc64le and aarch64.

The new oVirt version has several updates that improve the functionality and user experience of this open-source virtualization solution.

The contributions made to the new release were made by 46 developers within the community. This emphasizes the collaborative efforts made towards enhancing oVirt’s capabilities and addressing user feedback.

Key Features of oVirt 4.5.5

– Component Updates: The release features updates to several core components including:

• OTOPI: Now at version 1.10.4
• oVirt Ansible Collection: Updated to 3.2.0-1
• oVirt Engine Data Warehouse: Upgraded to 4.5.8
• oVirt Engine API Model: Version 4.6.0 is now available.

– High Availability Improvements: The Hosted Engine HA was updated to version 2.5.1, enhancing the resilience of hosted environments.

– API Enhancements: The release improves on the oVirt Engine API Metamodel (version 1.3.10) and the SDK (Python version 4.6.2), providing better tools for developers.

– Performance Monitoring Enhancements: Metrics collection has been upgraded to version 1.6.2, facilitating more effective virtual machine performance monitoring.

– Log Management Updates: The Log Collector is now at version 4.5.0, improving log data management across virtualized environments.

– Networking Enhancements with Open vSwitch: Updated integration with Open vSwitch versions 2.15-4 (el8) and 2.17-1 (el9) enhances networking capabilities within oVirt.

– Bug Fixes and Security Patches: This release addresses various bugs and security vulnerabilities, including:

• Fixes for issues related to VM import processes and disk configuration error handling.
• Security updates addressing vulnerabilities like CVE-2024-0822 involved disabling specific execution capabilities in GWT code.

oVirt as a Basis for Red Hat Virtualization and Oracle Linux Virtualization Manager

oVirt serves as the upstream open-source project for both Red Hat Virtualization (RHV) and Oracle Linux Virtualization Manager (OLVM). Its key role in these virtualization tools highlights its significance in the broader ecosystem of virtualization solutions. Both RHV and OLVM gets to benefit from the continuous development of oVirt. With the new release, both platforms can seamlessly  integrate these new features rapidly while maintaining stability and performance standards expected by enterprise users.

Red Hat Virtualization (RHV)

Red Hat Virtualisation (RHV) Red Hat Virtualisation, which is based on oVirt, delivers an enterprise-grade virtualization solution with additional Red Hat support services. It uses oVirt’s robust management capabilities and also provides features like enhanced security protocols, advanced monitoring tools, and dedicated support options tailored to enterprise customers. Thus, RHV is a suitable option for organizations seeking a reliable virtualization platform backed by professional support.

Oracle Linux Virtualization Manager (OLVM)

Similarly, OLVM is also based on oVirt technology but is tailored specifically for Oracle environments. It integrates seamlessly with Oracle’s suite of products, offering specialized features that cater to Oracle database workloads and applications. This allows OLVM to provide users with a familiar interface while simultaneously ensuring compatibility with Oracle’s ecosystem.

oVirt Backup and Recovery

Backup and recovery are critical components of any virtualization strategy. In an enterprise setting where data integrity and availability are crucial, robust backup solutions ensure organizations can recover quickly from data disasters and loss incidents. Let’s break down the different methods available in a way that’s easy to understand.

Understanding Backup Modes

When using Storware Backup and Recovery with oVirt 4 or later, you can choose from four different backup modes:

1. Disk Attachment:

   • Think of this as creating a digital copy of your VM. The VM’s metadata and disk files are stored separately.
   • Pros: Simple to understand.
   • Cons: Requires a proxy VM in each cluster, and incremental backups aren’t supported.

2. Disk Image Transfer:

   • This method creates a snapshot of your VM’s disks, including any changes made.
   • Pros: Supports incremental backups, and no proxy VM is needed.
   • Cons: Requires oVirt 4.2 or later.

3. SSH Transfer:

   • Data is transferred directly from the hypervisor using SSH.
   • Pros: Can be efficient, especially for smaller environments.
   • Cons: May require additional network configuration.

4. Change Block Tracking:

   • Only the parts of your disks that have changed are backed up, saving time and storage space.
   • Pros: Highly efficient for incremental backups.
   • Cons: Requires oVirt 4.4 or later with specific versions of Libvirt, qemu-kvm, and vdsm.

Learn more about available backup strategies for oVirt VMs

A Note on Best Practices

For the best possible backup experience, Red Hat recommends updating your oVirt environment to the latest version. This will ensure you have access to the most recent features and security updates.

Need more help?

If you have any questions or need further assistance, don’t hesitate to reach out to our team.

Conclusion

oVirt offers a versatile platform for virtualization management, and its backup and recovery capabilities play a crucial role in maintaining system integrity and availability. With features like full and incremental backups, application-consistent snapshots, Changed Block Tracking (CBT), and agentless backup, oVirt is a robust and scalable backup solution for organizations and businesses seeking reliable disaster recovery solutions.