Nmedia - Fotolia

Manage Learn to apply best practices and optimize your operations.

How a global file system interface eases data growing pains

If reaching the storage limits of a file server is an imminent threat at your organization, a global system may be the answer to your data growth woes.

In recent years, storage administrators have struggled with unprecedented data growth. Although this growth typically affects all areas of an organization, the storage of unstructured file data has proven to be particularly problematic. File systems, even those that are distributed across multiple servers, have inherent limits that may be easily reached. When an organization is confronted with these limits, a global file system may be of some assistance.

Historically, organizations have chosen from several courses of action when file system limits are reached. Options include the archival of cold data, the creation of a secondary file repository that also has the effect of creating a data silo or moving the data to object storage within a public cloud. While storage migrations to the cloud are often the preferred option, these migrations introduce their own challenges, such as unpredictable costs, poor performance and possibly even the need to recode applications.

Another option is to implement a global file system interface. Such interfaces provide software-defined storage capabilities that greatly simplify file storage and could eventually change the way that we buy, deploy and tier storage.

More than a consolidation option

Two of the leading global file system interface vendors are Panzura and Nasuni. Each one has its own way of doing things, but both provide ways for enterprises to begin phasing out legacy NAS in favor of storage that is more scalable and easier to manage.

On the surface, it might be tempting to dismiss this technology as being little more than data consolidation products. However, there are a few key capabilities that set global file storage tools apart from more mundane consolidation options, such as distributed file systems.

One of the primary ways Panzura and Nasuni differ from other file consolidation methods is in the way they store data. Typically, file servers store data using a file system, such as Network File System (NFS) or an NT file system. As previously explained, these file systems have limits. Rather than depending on a traditional file system, global file systems store data on object storage. Some only support cloud storage, while others support on-premises or hybrid storage.

Object storage does not use a traditional file system. Instead, it uses a flat storage structure in which files are assigned a simple identifier. Global file system interfaces act as a storage front end, which enables organizations to continue to access their data using NFS or something similar. The interface then translates the file system calls into RESTful API calls that are compatible with the back-end object storage. This approach enables unlimited file storage without the need for completely changing the way that users and applications access file data.

Benefits of global file storage

Global file storage interfaces also seek to improve performance by locally caching hot files.

One of the benefits to using a global file storage interface is file versioning. Versioning capabilities vary from one vendor to the next but can usually be configured to automatically retain multiple versions of files. This makes it possible to revert to previous file versions without having to perform backup and restore operations. Rollback may be possible on a per-file or per-volume basis. Furthermore, write once, read many capabilities can render previous file versions immutable. Such capabilities can offer organizations a huge advantage in the fight against ransomware.

Global file storage interfaces also seek to improve performance by locally caching hot files. Typically, an organization's unstructured data resides in an object storage repository where the data is deduplicated and encrypted. This data may also be georeplicated for the sake of redundancy or for providing faster access to the data from other regions. In either case, the most frequently and most recently accessed data can be cached locally to ensure optimal performance.

While global file system interfaces are probably best known as consolidation methods, they often bring to the table features such as data encryption -- both at rest and in transit -- deduplication and automatic versioning. On top of such capabilities, however, there may be other supplementary features that vary from one vendor to the next. For example, some vendors provide an interface that enables users to access their files through a web browser. Likewise, a vendor might also provide mobile applications for accessing file data. Another feature to watch for is native disaster recovery capabilities.

This was last published in September 2018

Dig Deeper on Storage management tools

Join the conversation

2 comments

Send me notifications when other members comment.

Please create a username to comment.

How do you confront file system limits in your organization?
Cancel
I do not think it is realistic for an organization to store in cloud its excess data from enterprise file server. That model has already been proven to be expensive. In early days of cloud storage people thought it to be a good idea but now they have realized that cloud storage like S3 can get very expensive too especially if you are dealing with large data storage. In addition the access latencies to access any data that is not local and stored in cloud can get very large depending on the geographic location of that data. As a result solutions like Panzura are practical only in certain use cases. The caching can only help if your data is in cache but you still have to take the hit for first time you get the non-local data. In addition protocols like NFS were never written for cloud storage. So if someone claims to have built a global file system using NFS front end to apps then I would seriously doubt that claim.
Cancel

-ADS BY GOOGLE

SearchDisasterRecovery

SearchDataBackup

SearchConvergedInfrastructure

Close