3 major errors to avoid when backing up virtual servers

There are many obvious benefits to migrating servers to virtual machines: hardware consolidation, performance management, and disaster recovery are much easier.

Even if there are some shortcomings, it is generated by the old infrastructure. Because most systems are designed to support only a certain number of physical servers, some special backup systems can be severely attacked. After using virtualization, your server count can grow exponentially, assuming that the use of resources is beyond control, and developing a solution before virtualization can cause backups to go out of control.

Major Error One: Laissez-faire

Agent-based backup is a common method for backing up servers. This method involves installing a backup agent on your own server and setting it up according to the backup server. Then you can back up these servers with almost the same configuration -- except for some application-specific differences -- to back up at specific times of the day or within a specific time frame.

Now use the same method as before and apply it to the virtual server. First, you need to run a dozen or more servers on a single physical host, and there are dozens or more backups on the same physical host. These backups and servers have the same CPU resources, the same access I/O operates and is likely to back up the same data at the same time. In this case, the virtual environment grows more rapidly than its equivalent physical environment, and you'll discover how to mine virtual host resources and backup infrastructure.

Major Error 2: Change everything without thinking about it

In order to deal with issues related to virtualized backup, virtualization and backup vendors will sell you a whole set of tools. But you need to compare these solutions carefully, especially since the price of some programs is tens of thousands or even hundreds of thousands, which is very expensive. In addition, these programs also have their own advantages and disadvantages.

A backup method is host-based backup, which backs up virtual machines into files. All virtual machines are essentially large files stored on disk, so backing up virtual machines without taking up resources allows you to back them up at once. This allows you to back up those virtual machines with a single backup task, reducing the number of agents installed and monitored. At the same time, it reduces the impact of resource-intensive functions such as proxy software compression.

You may have lost proxy granularity on each virtual machine because of this scenario. It's more difficult to fetch a single file from a particular backup, not to mention finding out which virtual machine is backed up in a large file. You lose file-level directory management, but most people rely on it for small-scale recovery.

In this case, backing up the data is simple, but it is complicated when it is restored. When dealing with features such as VMware's VMotion that allows virtual machines to be transferred on different hosts, you can be really a headache. This makes it very difficult to record virtual machines based on host backups.

When it comes to backup, you don't have to neglect the whole thing. The truth is that the agents running on each machine bring a lot of significant benefits. You are using an effective backup method that has been in use for many years. The agent allows you to manage granularity, and you also have a complete directory of which files are backed up and when. All the shortcomings have been mentioned before.

Comprehensive backup, also known as hot backup or offline backup, is a virtualization-specific backup technology. It stores a snapshot of the virtual machine and transmits it over a local area network (LAN) or Chaoleng's larger storage area network (the SAN transfers it to one of the backup proxy servers. The backup agent will track the contents of these snapshots and their sources. It will Manage everything.

Recovering a series of files requires the agent to load the hard drive and get the data. The downside here is that there are extra facilities. It requires a new hard drive, usually a disk-based backup, and as a whole backup. Part of the solution software. This solution is flexible but expensive, you need to start a new way to back up.

Major Error 3: Quick Decisions

for your organization It takes some work to make the right decision. In order to make the choices right, observe the current environment. Can traditional agents and even virtual servers on each server work for you? Deploy these on the virtual machine test platform. Measurement, CPU, network throughput and completion time. VMware's technology and Microsoft Hyper-V have their own count of collectable information So you can see the impact on the host machine.

How much data is stored in your backup environment? Do you have enough space? If you find that the proxy backup will have backup space or completion time Negative effects, please review your backup tasks, possibly using inefficient backup tasks. You don't need 500 copies of C:/Windows, you will always see unnecessary copy cases.

If the system Or network resources are the root of the problem, see if you can interleave tasks. On a single host, running all backup tasks at the same time will put pressure on the resources, but if the whole night is interlaced, you will see the use of those agents. Improvements.

If network bandwidth becomes a real bottleneck, consider preparing a separate virtual network for backup. It is common to use multiple virtual NICs and virtual local area networks (VLANs) inside the virtual switch. Consider reducing bandwidth pressure on different virtual LANs and physical NICs.

When considering host-based backup, look at you Whether the host agent is compatible with virtual machines. Backing up virtual machine files can cause inconsistencies, so there must be a proxy and task to manage the virtual machines. Know what service level agreements can be used for backup and recovery, because now restoring backup means hanging Load a virtual machine to get files, especially if those virtual disks are on tape. This is a two-step process and requires more disk space and time for the virtual machine to resume mounting.

A comprehensive backup will be a new alternative to your current solution. This is not a small project. If you remember the situation when setting up and fine-tuning the original backup infrastructure, you will understand that a new system will take some time to adapt. Remember that like other IT projects, you need new hardware, software, and configuration support – not a magic wand.

The Right Decision

Virtualization is very strong, it disrupts our original core system and support architecture. Backup is a basic service that allows you to make no mistakes. Take this opportunity to create a backup plan in a new virtual project so you can review the currently available methods. Making smart decisions will help you avoid major mistakes and give you a new understanding of old, reliable solutions.