VPs of Infrastructure and Operations are under constant pressure to more rapidly provision infrastructure to support applications, hit cost reduction targets, and maintain or improve application availability for network attached storage. To meet these goals, they are the directors of a carefully orchestrated dance of network, server, storage and datacenter operation teams.
Integrating Storage into Datacenter
Operations July 2007Executive Summary
VPs of Infrastructure and Operations are under constant pressure to more rapidly provision infrastructure to support applications, hit cost reduction targets, and maintain or improve application availability. To meet these goals, they are the directors of a carefully orchestrated dance of network, server, storage and datacenter operation teams.
Unfortunately, the state of storage management applications today has created the "black hole" of storage. Beyond the immediate storage team, information about the storage services delivered to an application is not easily available. SANscreen provides a storage service view useful to integrating storage operations into the balance of the infrastructure management team. The result is a decrease in application provision time, improved availability and a decreased capital cost.
The Black Hole of Storage
Storage is many times referred to as the black hole of the infrastructure environment. Many infrastructure team members go so far as to label it as the black hole of the team. This criticism has its roots in the independent nature of the storage team. As more and more servers take advantage of centralized storage, its prominence in the organization has grown. Unfortunately the applications used to manage storage have not developed in such a way to provide useful, actionable information to the balance of the infrastructure team. The communication for provisioning, application management and cost control between the storage teams and the rest of the IT ecosystem is still many times email and spreadsheets.
State of Storage Management Today
Today's storage management applications commonly called SRM applications do an adequate job of managing individual devices in the storage environment. Some applications manage switches, others manage arrays, still others attempt to track file level utilization for an application.
Some progress has been made to create unified screens of these devices. But all these applications manage devices. Storage is the first virtualized service deployed in the datacenter. It is at the cutting edge of IT as a service. Unfortunately, device level presentations of the environment miss the mark on this vision. Not only do they miss the mark, they are also unhelpful to the balance of the IT organization. IT application owners are not concerned about the configuration of switch 123 so much as they are concerned whether their application has the right level of recoverability. IT application owners are not concerned about volume masking as they are the total cost in resourcesto deliver their applications. Yet today's device centric storage managers reveal only this device centric information.
Today's storage management applications lack a service view of the storage infrastructure that is actionable by teams outside the storage environment. This lack of a service view impacts IT operations during the provisioning cycle, the operations cycle, and the chargeback cycle.
Storage Induced Challenges
Without a service view of the storage environment, cross functional tasks such as provisioning applications, maintaining applications, proving compliance, charging back, and predicting future capital costs are extremely difficult.
Provisioning Challenges During the provisioning cycle, the server team requests to the storage team tend to be made via an email or ticketing system. But there is no way to monitor the provisioning process as it occurs to gauge progress. The server teams literally wait until the storage team provides them the OK that storage is available.
In environments where the storage team is physically separated from the infrastructure and datacenter operations team, incorrect cabling and HBA placements can introduce errors can into the environment in the hand off between the storage and operations team.
Once the process is complete, there is no easy way to validate that the actual services requested have in fact been delivered. Application owners and the server teamare hard pressed to answer questions around:
· Is the tier of storage allocated correct?
· Do I have secure, redundant access to the storage arrays?
· What is the current recoverability of the storage?
· How much did this cost?
The net result of the hand offs, lack o... [download for more]
