Common Mistakes in the Storage Virtualization

[Lead Team]

Sponsored Links

Without careful planning, projects to go Virtualization of storage - or server infrastructures wrong. Many shortcomings in the consultation and planning are also covered in the implementation phase. This is especially true for cooperation with external service providers. In practice ten typical errors are observed.

The first error is from the service provider to advise the customer by asking for and accurately define the requirements and functions of the future system avoid. A typical misunderstanding between consultants and users, for example, the functioning of Alternate Pathing/multipathing. This is used by companies often understood as the application server can switch without interruption from one storage system to another. The customer is usually unaware that arises when certain manufacturers' products a certain switching time, since a script or even a manual intervention is necessary, and thus leads to downtime.

[Lead Team]

The second error that during the consultation and planning often occurs is that the customer does not meet its required I / O performance values is required. These are the correct dimensions of the storage server required. The sizing of the required components (FC and network cards or RAM) in the storage server is out of money, the lower limit is often made. A storage server should be sized so that it can intercept for the disaster case, a server load.

Error number three discussed is that the migration of existing data on the virtualized storage system ignored or inadequately with the customers. This leads to delays in the introduction.

A fourth common error: The advice provided by a service provider is indeed on the Backup and Disaster Recovery speaking opportunities, but remember this with the existing concept to reconcile.

[Lead Team]

Lack of consultation with the manufacturers or lack of review of the HCL (Hardware Compatibility List) is the fifth most common errors. But this is absolutely necessary, as certain storage systems has its own driver on the storage server need to be properly addressed. But there are storage servers that only a particular individual or driver may be used to communicate with the storage system or can. This one has to clarify in advance.

The sixth vulnerability affects the planning of the project. An appropriate concept for the project is often forgotten or cost reasons not even created. In this concept, all the important facts and features of the storage virtualization system are described. Furthermore, all key figures of the data to be migrated, a network diagram, a checklist of all the system data (versions and firmware versions) and a timetable be included.

During the installation of systems and components, error number seven occurs often in Release: The zoning of FC switch environment should be established so that the application server can not interfere in each other. For this reason, a 1-1 zoning would be favored. This is the device while a slightly higher cost, but has the advantage that the application server can interfere with each other.

[Lead Team]

The eighth error of the very committed to the implementation is also often is the lack of testing before the storage system is brought online, and the lack of documentation. This documentation must be made as large and well kept up, that perfect reconstruction of the systems in the event of a failure is possible.

Also in the maintenance of the systems are often made mistakes: monitoring and maintenance of the systems are usually operated so not how it would be necessary. Such systems have a critical business function and therefore require constant monitoring and care. This includes a regular review of disaster strategy and function. This test is - Error number nine - very often for reasons of time, as the systems feature an HA (highly available) have not carried out regularly.

Last but not least, tenth is the inadequate training of administrators mistake that many problems in this environment are attributed.