Busting 5 myths about data backups and hosting providers

Are data backups a necessity? Who is responsible for reserve copying, the provider or the client? Where should I store my data and how do I recover it? Certain hosting providers do not provide comprehensive to these questions, which is why there are plenty of myths concerning reserve copying. In this article, we will look into the 5 most popular misconceptions regarding backups.

Myth 1: I don’t need backups, I use expensive hardware

Using expensive Enterprise-class equipment is a good way to prevent data loss. However, many risks cannot be foreseen. Even high-quality hardware can have a faulty power supply unit and the risk of failure still exists in reliable data centers with multilevel fault-tolerance measures. These data centers use multiple reserve components for every element of the system; the infrastructure’s uptime does not depend on a single SSD drive or any other piece of equipment. This principle should be applied to backups as well, meaning that a user should not rely on the equipment alone. Using expensive hardware is an important preventative measure; however, it cannot be the only measure used by the company. Recreating corporate IT systems is a complicated and expensive task that will interrupt the company’s business processes. It is much easier to avoid this issue by copying the data of an Enterprise-class server into backup storage.

Myth 2: When I order hosting services, I get backups by default

In reality, backups are service, just like hosting or system integration; this is why «BaaS» stands for «Backup-as-a-Service». In most cases, reserve copying is provided for a fee on-demand, separately from the main service. The reason behind this is that creating and storing cloud backups takes time and resources. In order to create reserve copies, one must have space in the backup storage, physical space for the servers, specialized software and a system administrator to organize the process. Some providers do offer this service for free; however, this is rarely the case for large infrastructural projects. The more data needs to be copied, the larger the expenses. It is much easier to create a reserve copy of a single website than to set up backups for an entire corporate IT infrastructure.

Myth 3: One backup is enough to prevent data loss

The risk of failure and data loss is not limited to the company’s main infrastructure. The backup storage can also fail, just like any other server. In order to mitigate this risk, a company should have multiple reserve copies of the system. It is also advisable that one of the copies is stored on a remote platform; this way, failures in the main infrastructure will not affect the backups. Some providers offer multiple locations for backup storage. For example, SIM-Networks’ BaaS service has two options for clients: BaaS Local and BaaS Remote. BaaS Local stores incremental backups in the same data center where the client’s infrastructure is located; this increases recovery speed. BaaS Remote transfers reserve copies into storage located in another data center; the recovery time is slower than in BaaS Local, however, the risk of data loss is lowered. Having offsite backups in the cloud is a reliable measure to protect the company’s data from hardware failures or natural disasters.

Incremental backups only save the data that has been changed since the last reserve copy was created. This way, the copies are created quickly and take less space in the cloud backup storage.

Myth 4: Reserve copying is the same as fault tolerance

Reserve copying and fault tolerance are different concepts. Reserve copying is the creation of static copies of a certain array of data at a given moment. These copies are transferred to the data storage, ready to be used until they are obsolete. Fault tolerance is a set of measures intended to ensure that the hardware functions with no interruptions. For example, if one node happens to fail, another node will pick up its workloads and continue processing them. Reserve copying is used to save data in case the server or the cloud happens to fail; fault tolerance is a way of making sure that these failures will not cause infrastructure downtime.

Myth 5: The provider must recover my data regardless of the situation

Your provider will restore data from a reserve copy if it was lost due to a natural disaster, a mistake on the provider’s end or other unforeseen circumstances. However, if the data was lost due to the client’s actions, the provider is not responsible for recovering it. Let’s look into an example: a former system administrator of a US-based company got into the system after his employment was terminated and destroyed all the data he could. The company’s damages amounted to hundreds of thousands of dollars. This situation came about on the client’s side; the company’s management could have prevented it or changed the passwords beforehand. The provider cannot be held responsible for such events. In this particular case, the infrastructure vendor made an exception and helped recover the system; however, in most cases, such situations are beyond the provider’s area of responsibility according to the SLA.

What’s the best way to organize backups for my system?

As an infrastructure provider, we’re familiar with the subject of backup and recovery. Backups are a necessary condition for the stable functioning of any corporate system. This is why it’s vital to base your choice on the latest reliable information instead of widespread myths.

When choosing an infrastructural solution, look for providers that offer automatic reserve copying. For example, the BaaS cloud backup service for SIM-Cloud is a ready-for-use tool. If you need to create a backup, all you have to do is choose the data, set the backup depth and frequency; the data will be uploaded to the chosen data center automatically. We also offer 3 types of storage as an additional service for server rental: FTP, iSCSI and S3-compatible storage in the cloud.

Did you like the article?