Automated Storage Provisioning Solution for Application Growth
Data Center Challenges
Delivering storage provisioning services is one of the most complex and time-consuming activities in data centers today. Hundreds of new applications are introduced into SAN infrastructures each month and your IT staff is getting requests from various business units for additional storage allocations. This creates a rather large storage allocation request queue that is typically executed on weekends by IT staff members. The size of your team and the number of requests processed remains the same, but queued requests vary from week to week. IT staff members process as many allocation requests as possible given their time constraints. This type of operation is not scalable. Any requests that are not completed remains open till the next weekend or an available time slot.
The storage allocation process encompasses a significant number of labor intensive and time-consuming tasks. This is especially the case for allocating storage for new applications going live and existing applications growing in response to an up tick in business activity.
Different storage allocation methodologies (best practices) followed for storage provisioning and a host of tools are used to execute provisioning tasks. In the latter case, some companies leverage scripts, others leverage a manual ticket request process, and still others provision using point solutions. In any case, companies continue to be challenged with the response times. Here are examples of two:
- Over-provisioning; The IT staff provisions storage based on the application manger's request (which is typically more than is needed), or based on a forecast for a year's growth (estimate), or a plan for worst-case growth rates. In either case overprovisioning can be costly and lead to poor utilization rates. The result is unused capacity that sits on data center floors consuming power and floor space.
- "Just in time" provisioning; The IT staff provisions "just enough" storage to satisfy near term requirements with little to no room for peak conditions or spikes in activity. this is a risky practice that leads to downtime when additional storage is required (resize volumes, reconfigure applications).
Incipient Software - Solution
Incipient software automates the storage resource selection and storage provisioning process for large SAN environments significantly reducing the administrative steps to allocate new storage to applications. It takes advantage of the deep discovery capabilities it performed when discovering all of the storage arrays and switches within the SAN environment. It also leverages industry best practices that are embedded in its storage provisioning engine. Incipient software's discovery and provisioning capability are not found in any other network virtualization solutions. Other solutions require users to manually select resources and use multiple element managers to perform the provisioning tasks.
Pre-defined policies for setting, configuring and provisioning storage resources (storage arrays volumes) are made available to storage administrators to facilitate the entire end-to-end process. The following concepts are key features of Incipient software's automated provisioning capabilities:
- Deep Discovery
[+] Detail
Deep Discovery
performs probing and cataloging (discovery) of storage arrays and switches that are connected to an iNSP cluster. iNSP Storage Administrator discovers detailed configuration information about the layout of each storage array LUN, including detail "behind mappings" to the underlying physical disk drives. This information is used to make intelligent placement decisions when configuring iNSP virtual disks and consuming array LUNs. For example, both mirror sides of an iNSP volume would not be placed in the same failure domain (i.e. physical disk spindle).
- Discovery
- Switches, storage
- Incipient network volumes
- Deep array discovery support
- LUN to physical drive mapping
-
Storage Pools
[+] Detail
Storage Pools
allocate storage resources logically instead of physically. A storage pool is a collection of storage resources (bound array volumes) that contain all of the data for a specified set of iNSP network volumes. Storage pools can be grouped by customer (business unit), quality of service (QoS), or any combination that would be useful to an administrator. Storage pools simplify resource selection during provisioning operations by correlating storage types to appropriate server applications. When combined with iNSP authorization, administrators can effectively limit who can perform operations against the resources in a given storage pool.
-
Volume Classes
[+] Detail
Volume Classes
provide the rules used to select array resources (array volumes or LUNs) as well as configure
iNSP volumes, migrations or copies. Volume classes are provisioning templates, used to specify RAID attributes, number of paths, min. max. volume size and what storage Pools to select resources from.
Incipient Software - Payback (ROI)
Quantifiable savings are readily found in the following areas:
- Easy online and non-disruptive data migrations. Storage administrators no londer are required to over-provision to accommodate for future growth. This results in storage purchase delays (CAP-EX savings)
- Automate and simplify provisioning process, thereby reducing manual intervention; risk mitigation (OP-EX savings)
-
Increase storage utilization rates, leverage stranded capacity (silo) (CAP-EX savings)