Cloud Storage: FTP, iSCSI & NAS/NFS Backup System Solutions

Follow GigeNET on Twitter

Call Us Today: (800) 908-7115
Local Number: (224)-265-9165
Request a Call

Backup Solutions - FTP, iSCSI and NAS/NFS


This article explores the need for Backup, the important features of a backup system, as well as the different ways in which backups of systems are implemented.

Contents



A Need for Backup


While there is an instinctive understanding for the need for backup systems, we need to fully understand its purpose. This is because there is tremendous variety in backup solutions and only by understanding our needs can we make an informed choice.

The first and foremost need for a backup system is protection. Protection from disaster. Information and IT processes are so vital to businesses these days that a catastrophic event can cripple a company's operations. In that sense, backup is like insurance.

The second purpose of a backup is to restore mistakes made during normal operation. In case a few files are incorrectly changed or a certain individual element needs to be restored to an earlier state, backups provide the means for doing so.

Finally, there might be legal considerations. Banks for example are required to maintain records and a loss of them would not only destroy their operations, it would also leave them open to legal attack. These days we hear of ISP's being required to track users across the Internet. Data retention laws are becoming more stringent all the time.

So we see a wide variety of reasons for backups, and depending on the urgency and importance of these systems for your business, you will need to make an informed choice regarding how to go about it. In this article, we explain three techniques regarding the implementation of backup systems.

Features of Backup Systems


Backup systems need to be evaluated on the basis of several parameters. Any combination of these factors might be the "right" one for your firm depending on its needs.

Frequency of Backups


Backup Frequency is entirely dependent on the type of business that needs protection. Some business models are content to have backups relatively infrequently. This time period between backups can vary considerably. There are some businesses that can take a backup once a week or less and survive, whereas others need real time backups by the minute or even by the second.

Firms such as Banks and other businesses whose life blood is their data and IT infrastructure typically have more demanding backup schedules. To accomplish this, they select technologies that are able to provide stringent protection. These technologies are naturally more complex and demanding than the ones that are needed for more infrequent backups.

Amount of Storage Space and Other Resources


Any backup will consume storage space. The question that faces a firm is, how much is necessary. Storage space depends on two things - what is being backed up and how long it is retained. If the Recovery Point Objectives (RPO) require that a company should be able to restore it's data from a year ago then a very large amount of space is necessary as the backups will accumulate for a year.

On the other hand, a company can manage it's storage space for long periods of time if it realizes that it doesn't have to backup everything. As mentioned earlier, it all starts with an understanding of what the objectives are.

In addition to storage space, there are other resources that are consumed by a backup system. These are hardware as well as software resources. Also, a certain amount of labor is involved to run the system and the quantity increases with the complexity. Software to implement proper backup policies can become very expensive for commercial establishments and takes up a large proportion of the costs.

The trade offs between exhaustive backup systems and cost is one that each enterprise needs to take individually. It is important however, to design a system that is scalable in a way that allows it to increase the amount stored gracefully without significant further investment.

Security


Security is a much overlooked facet of Backup systems. However, the potential for security breaches causing tremendous damage is very high - especially considering that a backup contains ALL the data on the server. Most backup systems these days rely on the client/server architecture of the Internet since servers are typically stored away in data centers. Therefore a backup server usually initiates a connection with a client to initiate the transfer.

Imagine the following scenarios:

* An unauthorized server initiates a backup from a remote client. All the data is transferred to the server without checks
* An unauthorized client requests a backup from a server resulting in all the data being transferred to the client
* While the data is being transferred, it is intercepted by a "man in the middle" and stolen
* Backup storage devices such as disks and tapes are carried away by an unauthorized person

In any of these scenarios, the damage to a corporation can be devastating. Recently there have been some high profile media cases involving precisely this sort of data theft. If the business is a bank or any other institution carrying sensitive data, it might put hundreds of thousands of customers at risk of fraud.

Due to this, it is very important for a backup infrastructure to put in place software, hardware and personnel policies that are enforced strictly. The amount of security your business needs has to be determined by you taking into consideration the costs as well as the type of business that is carried out over the IT network.

Later in this article, we take a look at each of the systems that can be used for backup and analyze the various security implications that attend them. We also look at how the security can be improved for each system.

Performance


Whenever you backup a client machine, there is a performance cost. This cost can be measured in terms of how well the client responds to queries while the backup is in progress, or in terms of how much bandwidth is available over the network at the time of backup. In case the process takes a lot of time, IT administrators have to ensure that the whole system doesn't slow down to a crawl. In case the data is being transferred over the Internet, special attention needs to be given to how it will affect the data transfer costs.

Several technologies exist that allow backups to be faster. Compression for example, reduces the amount of data that needs to be transferred. With the emergence of Continuous Data Protection (CDP), only the data that reflects changes to files is transferred, removing the need for redundant data transfer.

In order to minimize the impact, technologies such as temporary disk caches to hold changes can be implemented until the resources are available to send them to the backup.

Yet other administrators try and work around the performance impacts in such a way that their systems have no other load at that time - either by taking the server offline for a short time or scheduling the backup at night or another time when experience tells them the load will be low.

The different technologies for backup examined in this article have various performance impacts on both the client as well as the server. We will examine each of them in turn. Once more, as with all other parameters, there is a trade off between cost and benefits and each business needs to understand what their objectives are and then find a compromise.

Cost


There are several components of cost in Backup systems. They can be neatly classified into hardware, software, bandwidth, and administration costs. Hardware costs will include servers which initiate backups and associated infrastructure, storage media (either tapes or drives) and any other network equipment.

The software costs refer to the price of the Backup software that you purchase. For large corporations, this can be a sizable amount and in fact, can make up the bulk of the pricing. To a large extent, it depends on how complex your infrastructure is and what your Recovery Point Objectives are.

Bandwidth costs can soar if you take regular backups over the Internet. Most datacenters however, allow unlimited transfer between servers of the same group, and so there are tremendous potentials for cost savings here.

Finally, man power and administration costs are an important factor. Not only do they include the salaries of the administrators who manage the backups, they must also include costs for maintaining and storing the backups and security procedures for the same.

Reporting


An extremely important facet of Backup systems is monitoring them. Backup processes can be so complex that anything can go wrong. It's important to have a finger on the pulse of the operation.

Conversely, you don't want a system that tell you everything. Information overload is just as bad as information paucity. So we'd like a system that tells us what we need to know - no more and no less. Also, we need a system that tells the correct person. Meaning that everyone who is in the loop, gets to know what he or she needs to know.

The way the reports are generated are important as well. You can have reports sent to you by mail, or you can access them on the server. For maximum accessibility, the reports should ideally be available over the Internet via a web browser through a private VPN connection. This way, you don't need to be tied down to specific machines with software on them to view information.

Reporting can be set up to alert the responsible administrators when specific events happen. Perhaps you would like to know when a certain error occurs or the amount of data being backed up exceeds a certain limit. All these options allow you to stay on top of what is happening and fine tune your backup procedures to more accurately suit your organization.

Backup Solutions


In this section, we examine three primary technologies for backups systems. These are FTP, ISCSI, and NAS/NFS. The parameters for comparison will be the ones listed above. Each technology has it's own "gotchas" and advantages. You will need to make the choice based on the needs of your business and your budget.

FTP


FTP Backup Software refers to a methodology of backup that relies on the offline storage model. In this model, the data that is backup up is not stored in the same location. There are several benefits to this model, the most important being that a data center wide disaster can be survived since the data is still in another location. There are disadvantages too, not the least of which is bandwidth.

FTP is the simplest method of transferring files from one server to another. As such, it is ideal for businesses or individuals who need to backup their data in a fast and easy way. In this section, we take a look at some of the features of FTP Backup.

Advantages of FTP Backup


The primary advantage of FTP backup is simplicity and (short term) costs. FTP Backups backups very easy to set up and with proper FTP Backup Software, configuration can be a breeze.

FTP backup is ideal if you need to protect a few important files and documents from disaster. This is true if you are an individual and you need a backup system for your own personal needs.

For more sensitive data, security can be improved by using SFTP and encryption technologies which are also easy to set up. For businesses, all that is needed is for an FTP server to be in place and a good FTP backup software.

Disadvantages of FTP Backup


Unfortunately, along with the simplicity of FTP come many disadvantages. First of all, FTP is not very fast. Each file requires a new TCP connection to be made. If there are many files that need to be backed up, this can go on for hours. Not very efficient at all. You might have Firewall problems because of this as well.

Secondly, FTP is not at all secure. Your username and password are sent in plain text to the FTP server and can be seen by anyone snooping on the network. In addition, FTP files are by default not encrypted when transferred. The makes your data easily visible online to snoopers.

Another factor is the cost of storage. For small storage requirements, FTP backup can be cost effective, but when large amounts of data are involved, you end up paying much more per MB of storage on a remote server. Thus FTP is practically useless for large enterprises.

FTP Backup Software


There is a lot of FTP Backup Software that works well for backing up small amounts of data. Such software can perform additional functions that ease the entire process such as encrypting content and scheduling backups when the client is not in use. In addition, the software can also take care of burning the backup up data to other media such as a disc.

FTP Backup Software can also perform advanced functions such as incremental backups. Some filetypes such as registry files may not be supported however.

iSCSI


Understanding iSCSI


ISCSI stands for Internet SCSI. In order to understand it, we must first understand what SCSI is. Simply put, SCSI (Small Computer Systems Interface) is a standardized way for devices to talk to each other. It is important because it allows for extremely high data exchanges between devices. This is ideal for backups systems since during backups, large amounts of data are transferred between storage systems.

Previously, SCSI could only be used between devices at most a few meters apart. The need has arisen however, for a backup server to be located far away from the client - perhaps in another datacenter in another country even. In order to allow SCSI commands to span such distances, a new protocol was created that would allow SCSI devices to talk to each other over any TCP connection like Ethernet. The new protocol was named ISCSI.

Using ISCSI, devices can transfer data between each other at the highest speeds that the bandwidth will allow. In fact, the emergence of ISCSI has led to a new paradigm in the way data stores are managed leading to the widespread prevalence of the Storage Area Network (SAN)

Storage Area Networks


ISCSI is an implementation of the Storage Area Network (SAN) concept. Under this paradigm, all servers share the total storage space between them. This leads to much better overall utilization of disk space. It's essentially the virtualization of storage devices.

The challenge arises when the storage devices are placed far away from the servers that need them. This is because there are several requirements that must be met - such as blazing speed. Also, the disks must appear to the servers to be locally attached. This means that the servers must have block level access to the storage devices.

This is in contrast to a system like FTP where the client and the server talk to each other on a software level. The client has no knowledge of the actual physical disks that the server is using. For efficient data transfer however, servers need to be able to address the disks directly.

As far as backup is concerned, SANs allow backups to be much faster than an FTP backup system. Administrators can decide whether to have a dedicated network infrastructure for their backup systems or whether they can let it piggyback on their existing systems.

For a long time, the primary method of implementing SAN systems was through Fiberchannel. This is a combination of hardware and protocols that deliver extremely fast data transfer. Much higher than the same over Ethernet. However, in most applications, it's not the width of the pipe that is the bottleneck, but the rate at which data can be fed into and taken out of the pipe. Therefore, most applications do not use the incredible transfer rates that it is theoretically possible to achieve using Fiber channel.

iSCSI uses Ethernet and in real world applications, provides extremely comparable performance when placed side by side with Fiber channel. This, along with the lower hardware costs and already existing infrastructure and administration has propelled iSCSI to a position where it is now on a par with Fiber channel in terms of adoption.

iSCSI Benefits


iSCSI runs on Ethernet and that is a huge advantage for several reasons. In the first place, all companies already have ethernet infrastructure built into their IT systems. Ethernet is useful for so many purposes including connecting to the Internet, that it is ubiquitous. This means that firms do not need to purchase additional hardware to implement iSCSI backup systems.

In addition, administrators are already familiar with the way ethernet protocols work. They know the routers, the ports, and how to configure TCP IP using the Operating System they are familiar with. This means less training for staff and greater ease of implementation. It's much more difficult to find and train staff who are going to manage Fiber channel.

Ethernet is also cheap. Much cheaper than Fiber channel since there are more people using it, and it has a lesser bandwidth. However, as seen above, that does not matter so much.

Finally, Ethernet can be scaled easily. There are literally no limits to how far the actual disks can be stored from the main server. This is due to the inherent advantages of TCP IP over Ethernet. Any system that is connected to the Internet can be connected using iSCSI over Ethernet. This allows for a multitude of architectures that are not otherwise possible. For example, you can have mirrored backups in distant datacenters.

NAS/NFS


Overview of NAS


NAS or Network Attached Storage is an architecture as well as a technology for data storage. It has it's uses in Backup systems just like SAN does. However, there are some significant differences as we shall see.

Differences from SAN


Unlike SAN, where the disks appear to the operating system on the server to be locally attached, NAS presents a file structure as an interface. In this respect, it is closer to a file server than anything else. Only that NAS is a much more efficient and stripped down version of a file server.

The way it works is simple. There is one central storage location containing the required number of storage devices. This location is connected to all the other systems in the network via simple Ethernet connections. Unlike SAN where by and large fiber channel is used for connection, NAS makes use of the existing infrastructure.

In SAN as we saw, block level transactions can be made. This is a direct consequence of the disk being viewed as a local attachment. In NAS however, file level commands are sent and there is little control over the actual blocks. This has both advantages and disadvantages as we see below.

iSCSI vs NAS


iSCSI and NAS share a number of similarities at first glance. They're both used to scale up storage over a network and can be used for several different purposes of backup. However, the underlying architecture causes some limitations when it comes to which applications can be used on NAS and iSCSI.

iSCSI networks allow the underlying disks to be viewed at a block level instead of at a file level, giving the impression that the disks are locally attached to the server as under Direct Attached Storage (DAS). Applications that rely on block level information can therefore be run on iSCSI SANs. These are applications that utilize disk block information for maximum efficiency. Examples include database management systems as well as various backup systems.

NAS systems on the other hand are used for applications which interact with the hard drives through the buffer of the underlying operating system. Meaning that they merely deal with files and folders. Some backup systems are comfortable working with this sort of interface and will function well.

Ultimately the choice to use iSCSI or NAS depends on what applications are being used. Since both use the Ethernet architecture, speeds of data transfer are more or less the same - though efficiency is slightly higher in iSCSI due to less processing requirements and fewer layers of protocol. The cost differential between the two systems is getting lower and lower as time goes by.

GigeNET and Storage Solutions


GigeNET provides advanced iSCSI SAN storage for complete data redundancy and backup. With our technicians taking away the hassle of configuring and setting up the hardware, you can have complete peace of mind as you manage your SAN from a web administration panel.

Complete storage flexibility and ease of use combine to create a compelling offering at extremely low prices going down to $0.75 per GB of storage! Order your iSCSI storage at our website to start!


Have a Question or Need Help?

Features at a Glance

  • Redundant Locations - LA Cloud Servers and Chicago Cloud Servers now available!
  • Fully Redundant, Self Healing SAN's CPU utilization based on the amount memory assigned to a machine.
  • Xen Hypervisor using Citrix Xenserver Technology.
  • A friendly, reliable control panel that offers usability second to none.
  • Multiple elastic public IP addresses.
  • Support for multiple linux distributions.
  • No reselling or over commission of resources.
  • Direct KVM console access that can be used to recover a file system failure or malformed network configuration.
  • Completely transparent private networking with free unmetered bandwidth.
  • All cloud nodes are based on 64-bit architectures.
  • Ability for Virtual Machines to use 32-bit or 64-bit architectures.
  • Full Windows Server 2003 and 2008 Support for all Cloud Servers!

Cloud Products

Keep up to date with Cloud Computing

Drop1 Button1 BG1 BG2 BG3