The petabyte era is increasingly presenting storage challenges to enterprises and other organizations:unstructured data is growing exponentially, and demands for availability, performance, and agility are rising. In addition, increasingly complex tasks are expected to be performed more and more efficient and with low maintenance. MSP is using technology that is radically easy to manage and scale: The Qumulo File Fabric (QF2) manages trillions of files and thousands of petabytes with minimal overhead.
Virtually unlimited scaling
Start with a cluster of four nodes and about 100 terabytes and scale up to nine exabytes. Qumulo supports trillions of files per cluster. This gives you instant access to all your data and metadata, making it fast and immediately usable. Storage size and performance scale linearly, as each storage node brings storage, compute and network capacity to the table. With conventional storage systems, a high number of files typically causes that recurring processes such as tree walks* can no longer be performed in an acceptable period of time. Directory structure and metadata become Big Data. Qumulo makes tree walks obsolete. The file system is based on a new approach that uses parallel, distributed algorithms. These principles of modern databases enable unprecedented scaling. Good to know: Qumulo is cost-effective starting at 100 terabytes.
represent operations in a file system that is structured like an inverted tree structure. The folders form a descending directory tree. Each node in this structure lists the files it contains, as well as data about the files and the subfolders they contain.
Intelligent caching makes hybrid systems consisting of SSD (Solid State Drive) and HDD (Hard Disk Drive) highly performant. All random write operations of the hybrid systems take place on SSD. A single cluster achieves more than one million IOPS (Input/Output Operations per Second), 333GB/s read and 200GB/s write.
Simple and efficient
Another advantage of Qumulo is its minimal management overhead: All data in a cluster are located in one namespace. Managing media types, disk groups, RAID and LUNs is a thing of the past. You therefore avoid data silos and use 100 percent of the available capacity. Even a fully utilized cluster is still capable of deleting data, enforcing quotas, and performing upgrades. This 100 percent usability and the small block size of 4K make Qumulo unbeatably efficient. Especially in environments with a high number of small files, Qumulo saves a lot of capacity.
Convergence is everything - Silos are the past
Agile product development and rapid change demand integrated solutions that adapt to the dynamics of your business. Qumulo takes an API-first approach, allowing you to address all storage operations via REST API from your software and integrate them directly into your
workflows. This creates a high level of automation and eliminates time-consuming tasks for storage admins.
Real-time monitoring and analysis
Get an always up-to-date insight into your data structure and all movements and performance data in real time. A clear Graphical User Interface (GUI) always shows you up-to-date data on IOPS, latency, bandwidths and sizes based on clusters, nodes, shares or users.