New version of Veeam Backup & Replication v8 (paid and free editions)

Last week, a new version of Veeam Backup & Replication v8 was released, and in this article I will describe the new functionality that appeared in it, as well as what is available from this new functionality in the free edition of the product.

NetApp hardware snapshots support

In Veeam Backup & Replication v8.0, it became possible to create backups and replicas of virtual machines based on hardware snapshots of NetApp storage systems. A similar approach has already been implemented in previous versions for such sorts of storage systems like HP StoreServ (3PAR), HP StoreVirtual (LeftHand / P4000) and HP StoreVirtual VSA.

Let me remind you of the advantages of hardware storage snapshots. The technology of disk hardware snapshots has long been designated by experts as a technology that can drastically reduce RPO (the maximum time during which new or changed data remains unprotected from failure, that is, not duplicated in backup). The essence of the technology lies in the fact that the storage system independently, with adjustable frequency, creates snapshots of its disks. Since storage developers are better off knowing how to build their disks at a low level, it is obvious that they can also take snapshots of such disks with the maximum level of efficiency and performance — that is, with the least impact on the productive system. Therefore, hardware snapshots are a good source of backups.

However, by themselves, hardware snapshots cannot be considered as a reliable way to protect data from failures, primarily because snapshots are located on the same storage system that they “reserve”. In the event of a failure in the storage system, both the original data and their snapshots are likely to fail. In order to ensure proper reliability of data protection against failures, it is necessary to transfer backup copies of data to another location relative to their original storage location.
')
In addition, snapshots have a number of disadvantages:

hardware snapshots use the usable capacity of the original disk , and their total maximum size is dictated not only by the requirements of data protection, but by the “counter” requirements of the business applications of the productive system. Therefore, the second urgent task is the timely removal of old snapshots. Moreover, due to the peculiarities of the algorithms used, the presence of a large number of old snapshots in some cases can slow down or even lead to failures in the operation of some applications whose data is located on the storage system. Veeam Backup & Replication allows you to simply solve this problem, since it “keeps snapshots” on the storage only until its data is transferred to the backup, after which the snapshot is automatically deleted.
the recovery procedure from the hardware snapshot is not granular (if necessary, restore a separate file to the storage, as a result of recovering the disk from the hardware snapshot, all storage is completely “moved to the past” and the last changes on the disk are lost),
hardware snapshot can be “crash-consistent” - since the storage system may have limitations for application backup backup, and it can create a backup without first resetting the application buffers in memory to disk, since in general the storage equipment is not has complete information about the nature of the data (about the type of operating system, about the type of disks: virtual or physical disks, about the type of applications, etc.) stored on it.

As a consequence of the above, it is quite reasonable to combine the functionality of the backup product and the functionality of hardware snapshots embedded in common storage systems, since they complement each other well.

More information about NetApp support can be found in our article on Habré "Support for hardware snapshots of NetApp storage systems in Veeam Backup & Replication v8"

EMC Data Domain Boost Support

Thanks to the integration of Veeam Backup & Replication v8 and EMC Data Domain Boost, the process of building full synthetic backups has been greatly accelerated (when each new incremental or differential backup is immediately used to rebuild the full copy of data in the repository, which significantly reduces recovery time) and the backup process copying in general. This became possible due to the advanced EMC Data Domain Boost deduplication functionality:

first, the backup process is accelerated by using deduplication technology on both sides when transferring data over the network (the amount of physically transmitted data is significantly reduced),
secondly, the construction of full synthetic backups in the storage with the built-in deduplication enabled is faster, since the data on the disk is moved logically, but not physically.

For example, in one experiment we conducted to create a full synthetic copy of a test virtual machine, the time taken to complete a task was less than 11.5 minutes when running from EMC Data Domain Boost, and almost 3 hours when starting a task with deduplication turned off. The difference in speed was about 15 times!

Of course, each specific virtual infrastructure is unique in its characteristics, and the gain from deduplication on a particular virtual machine will strongly depend on the nature and intensity of data changes on its disk.

And even if you prefer the usual full backup to synthetic, you will get the gain due to deduplication of network traffic: data blocks that have already been transferred to the storage side will not be retransmitted through the network in further full copies. This is clearly seen in the two pictures below (they reflect the network load when building the first incremental copy). In the picture on the left, traffic deduplication is enabled (EMC Data Domain Boost), and on the right - disabled.

What is better: synthetic or regular full backup?

Since EMC Data Domain has excellent built-in deduplication, creating multiple full copies does not overrun disk space, since repeated blocks (and most of them) of different full backups are stored in a single copy. At the same time, having a full backup minimizes data recovery time after a failure (RTO), since there is no need to look for the latest version of the data from multiple incremental backups.

From the point of view of restoring the last state of the system, the synthetic copy and the regular full copy are the same. However, there may be situations when you need to roll back more time into the past (say, when a damaged file got into the backup, or the backup itself is damaged) - and then the full-copy chain option has an undeniable advantage over the synthetic copy.

Therefore, answering the question “What is better?”, It can be argued that in the case of EMC Data Domain with deduplication enabled, one can give preference to the chain of full backups (unless, of course, there are no contraindications in terms of the load on the virtual infrastructure, since consumes significantly more resources than incremental).

Configuring EMC Data Domain Boost in Veeam Backup & Replication v8

In the Create Repository Wizard, the step of configuring EMC Data Domain as the target storage will appear.

When recording backups to EMC Data Domain, it is possible to use the “Fiber Channel” option, which allows you to unload the LAN.

New to the Veeam Explorers tool series

A series of tools for granular restoration of application objects Veeam Explorers is expanding again - the new version includes Veeam Explorer for Microsoft Active Directory and Veeam Explorer for Microsoft SQL Server, as well as the new features of Veeam Explorer for Microsoft Exchange. The latter is now able to restore objects after the so-called hard deletion. That is, for example, it is now possible to recover those emails that the user deleted “without the possibility of recovery” (via SHIFT + DEL in Outlook).

Database administrators will be able to restore the database from a backup of the SQL server without raising it, and the Active Directory administrator to restore the computer account from the OU will only need to specify which backup copy of the domain controller to use, or enter the path to the NTDS database .DIT. I note especially that the tools of the Veeam Explorers series are available in all editions, including the free “Free Edition”.

You can learn more about the new tools of the Veeam Explorers family and their capabilities in the article on Habré “What's new in Veeam Backup & Replication v8: Veeam Explorer for Active Directory and Veeam Explorer for Microsoft SQL Server” .

Cloud Storage Support with Veeam Cloud Connect

Veeam Cloud Connect is a new component of the Veeam Backup & Replication solution that is expected to help users quickly and securely, using SSL, transfer data to service providers without resorting to VPN. The opportunities provided by service providers and Veeam Cloud Connect eliminate the need for additional expenses for creating a remote site and dedicated repositories for storing backup copies.

New replication features

The following new replication features are available in Veeam Backup & Replication v8:

WAN-acceleration in the process of replication (previously, WAN-acceleration worked only in the mode of creating and transferring backup copies, and was not used in the process of replication);
scheduling a switch to a replica — you can create a script for the switch process, that is, group the virtual machines that will participate in the switch, set their launch order, set the start timer and other parameters;
scheduled replicas switching (not to be confused with scheduling, that is, with script creation) - now you can set the time when the source virtual machine should be turned off and the replica turned on; This is convenient, for example, for service.

More information about new replication features can be found in the article on Habré "New features of Veeam Backup & Replication v8.0: improved replication . "

Automatic control of the intensity of read / write operations

Veeam Backup & Replication v8.0 implements a method of automatic control over the intensity of read / write operations Backup I / O Control. Backup is a resource-intensive process that can absorb a significant portion of the resources of a productive system, so that its users can notice it, which will reduce their productivity. Therefore, in the version of v8 Veeam, B & R can self-limit itself — all you have to do is set a valid value for datastore latency (IOPS) and set the optimum combination of parameters for your infrastructure that will ensure maximum performance of the backup process at a given company’s workload.

More information about the automatic control mechanism is described in the article on Habré "New features of Veeam Backup & Replication v8.0: automatic control of the intensity of read / write operations"

What's new in the free edition

The free version of the product allows you to implement basic backup scenarios that are typical for non-critical network resources, for home users, or for very small companies. A full review of the features of the free version can be found in this article on Habré " Veeam Backup Free Edition v7 - review of the new version ." In general, the free version has several limitations, for example:

Only a full backup, only individual virtual machines and only in the manual start mode, without a schedule;
There is no way to run back-up virtual machines directly from the backup repository;
No API for scripting.

The table below shows what new functionality is available in the free edition of Veeam Backup v8.0 Free Edition . A full description of all the features of the free edition can be found here on the site.
The above-described new functionality, which is included in the free edition of the “Free Edition”:

New functionality	Available in Free Edition
Veeam Explorer for Microsoft Active Directory	Yes*
Veeam Explorer for Microsoft SQL	Yes*
NetApp hardware snapshot support	Not
EMC Data Domain Boost Support	Not
Replication	Not
Automatic control of the intensity of read / write operations	Not

* For restrictions on the free version, see here .

Additional links

Source: https://habr.com/ru/post/242587/

All Articles