Technology data center continues to grow each year. Then, how these trends in 2011? Consider the prediction Hu Yoshida, Chief Technology Officer (CTO), Hitachi Data Systems.
According to Hu, in 2011 will be colored with 10 main storage trends surrounding the transformation of the data center. Here penjabarannya as quoted on Tuesday (01/11/2011):
1. Adoption of storage virtualization and dynamic provisioning faster. Both the foundation for the implementation of cloud computing and dynamic data center with high availability. Storage virtualization, virtualization of external storage arrays enable the migration from one array to another array without any interruption. Meanwhile, dynamic provisioning enables storage is provided in minutes.
2. The integration of server and storage virtualization more closely. This integration is necessary to increase the adoption of virtualized data center. Today, server virtualization has matured beyond consolidation cost reductions from the print server, file, test and development to support the application server tier 1. Virtualization servers require the integration of virtualization storage array that can reduce the 'bottleneck' I / O software such as SCSI, so they can meet the needs of enterprise-class applications.
3. Adoption of virtual tiering for lifecycle management. Currently, tiering is able to allocate virtual volumes to a single pool of storage (storage group) that contains the performance, cost and number of tiers of storage. Virtual tiering also has the intelligence to move the components of that volume to another tier based on the number of access.
Users do not need to classify a volume and allocate it to a single tier storage. Users also do not need to raise or lower the volume to a certain tier based on the activity. Dynamic Virtual tiering or tiering will do it automatically without the need to inter-tier classification or transfer.
4. Time to use the SSD (Solid State Drive) in a virtual configuration tier. SSD can improve performance while lowering the cost of a virtual configuration tier. Currently, 80% more volume classified as inactive, only a small number of SSDs that are needed in Tier 1 to serve the active components from a single volume.
Most of the volume can use SAS or SATA drives are cheaper. Multi-tier storage pool that has a little SSD and a lot of SAS or SATA drives will be cheaper than a pool of SAS drives with a total capacity of the same and give IOP of four to five times as much.
5. Adoption of Serial Attached SCSI (SAS) for enterprise storage systems. Unlike Fibre Channel (FC) loop is used to support FC drives on the old storage systems, SAS is a protocol point-to-point. FC loop requires every drive in the loop to help access to the loop, which resulted in inefficiencies. If you drive faster - such as SSD drives - connected to loop it, will spend the loop so that other drives do not miss out access.
Most SAS drives currently 6Gbps speed and most of the FC loop is 4Gbps. Thus, SAS has faster performance with the access point-to-point, making them easier to find the drive that failed compared to FC loops. SAS is also compatible with SATA. The only difference is its port, SAS has a dual port SATA ports while the single.
6. Drive SFF (Small Form Factor), will be more easily found. Mainly because of the power and cooling efficiency SFF owned. SFF devices are 2.5 inch drives which consume power around 6 to 8 watts, compared to drive Large Form Factor (LFF) 3.5 inch which takes 12 to 15 watts. Thus, SFF cut dramatically the power and cooling, plus also savings from the use of the room.
Some vendors package the 24 disks in a drawer SFF 2U-high 33.5 inches wide. Hitachi AMS package and Virtual Storage Platform (VSP) to be more compact. One drawer AMS has 48 drive 3U high by 24 inches wide. This drawer can be pulled for maintenance by 48 disk is still spinning.
7. Cloud accepted as one of the infrastructure model. Cloud computing will begin to be accepted as one of the concepts tested. A number of foundation to the Clouds will encourage adoption by providing a number of device management and orchestration layer that provides transparency throughout the process.
8. Convergence data center soared. Convergence server infrastructure, storage and networking will simplify and accelerate application deployment. The use of servers, hypervisor, storage and network virtualization provides a way for an open platform that ensures investment protection and choice to users.
9. Applications increasingly transparent in storage virtualization or application will require a Cloud infrastructure. Without this transparency, the user application can not know whether the target Service Level Agreement (SLA) has been reached, how to know their billing mechanisms, planning and utilization of health level of its infrastructure.
Management software can provide a dashboard application for business units that use them in accordance with the target SLA. Dashboard shows the status of SLA, the allocation of disks, RAID type, port storage and the level of health infrastructure used, and capacity usage.
10. Managed services (managed services) Remote start is available.
Remote Managed Services will reduce the burden of monitoring, reporting, warning and management of IT operations that prevent switching to new technology. For more than 10 years, the mandate given to IT is to do more with less so that the staff was busy with the maintenance of the same technology. To transform the data center, IT staff must find time for training, planning and execution.
A group of IT experts who operate outside the Service Operations Center uses remote management devices can increase this capacity to the various installations at a cost of more acceptable and encourage faster return on investment.
According to Hu, in 2011 will be colored with 10 main storage trends surrounding the transformation of the data center. Here penjabarannya as quoted on Tuesday (01/11/2011):
1. Adoption of storage virtualization and dynamic provisioning faster. Both the foundation for the implementation of cloud computing and dynamic data center with high availability. Storage virtualization, virtualization of external storage arrays enable the migration from one array to another array without any interruption. Meanwhile, dynamic provisioning enables storage is provided in minutes.
2. The integration of server and storage virtualization more closely. This integration is necessary to increase the adoption of virtualized data center. Today, server virtualization has matured beyond consolidation cost reductions from the print server, file, test and development to support the application server tier 1. Virtualization servers require the integration of virtualization storage array that can reduce the 'bottleneck' I / O software such as SCSI, so they can meet the needs of enterprise-class applications.
3. Adoption of virtual tiering for lifecycle management. Currently, tiering is able to allocate virtual volumes to a single pool of storage (storage group) that contains the performance, cost and number of tiers of storage. Virtual tiering also has the intelligence to move the components of that volume to another tier based on the number of access.
Users do not need to classify a volume and allocate it to a single tier storage. Users also do not need to raise or lower the volume to a certain tier based on the activity. Dynamic Virtual tiering or tiering will do it automatically without the need to inter-tier classification or transfer.
4. Time to use the SSD (Solid State Drive) in a virtual configuration tier. SSD can improve performance while lowering the cost of a virtual configuration tier. Currently, 80% more volume classified as inactive, only a small number of SSDs that are needed in Tier 1 to serve the active components from a single volume.
Most of the volume can use SAS or SATA drives are cheaper. Multi-tier storage pool that has a little SSD and a lot of SAS or SATA drives will be cheaper than a pool of SAS drives with a total capacity of the same and give IOP of four to five times as much.
5. Adoption of Serial Attached SCSI (SAS) for enterprise storage systems. Unlike Fibre Channel (FC) loop is used to support FC drives on the old storage systems, SAS is a protocol point-to-point. FC loop requires every drive in the loop to help access to the loop, which resulted in inefficiencies. If you drive faster - such as SSD drives - connected to loop it, will spend the loop so that other drives do not miss out access.
Most SAS drives currently 6Gbps speed and most of the FC loop is 4Gbps. Thus, SAS has faster performance with the access point-to-point, making them easier to find the drive that failed compared to FC loops. SAS is also compatible with SATA. The only difference is its port, SAS has a dual port SATA ports while the single.
6. Drive SFF (Small Form Factor), will be more easily found. Mainly because of the power and cooling efficiency SFF owned. SFF devices are 2.5 inch drives which consume power around 6 to 8 watts, compared to drive Large Form Factor (LFF) 3.5 inch which takes 12 to 15 watts. Thus, SFF cut dramatically the power and cooling, plus also savings from the use of the room.
Some vendors package the 24 disks in a drawer SFF 2U-high 33.5 inches wide. Hitachi AMS package and Virtual Storage Platform (VSP) to be more compact. One drawer AMS has 48 drive 3U high by 24 inches wide. This drawer can be pulled for maintenance by 48 disk is still spinning.
7. Cloud accepted as one of the infrastructure model. Cloud computing will begin to be accepted as one of the concepts tested. A number of foundation to the Clouds will encourage adoption by providing a number of device management and orchestration layer that provides transparency throughout the process.
8. Convergence data center soared. Convergence server infrastructure, storage and networking will simplify and accelerate application deployment. The use of servers, hypervisor, storage and network virtualization provides a way for an open platform that ensures investment protection and choice to users.
9. Applications increasingly transparent in storage virtualization or application will require a Cloud infrastructure. Without this transparency, the user application can not know whether the target Service Level Agreement (SLA) has been reached, how to know their billing mechanisms, planning and utilization of health level of its infrastructure.
Management software can provide a dashboard application for business units that use them in accordance with the target SLA. Dashboard shows the status of SLA, the allocation of disks, RAID type, port storage and the level of health infrastructure used, and capacity usage.
10. Managed services (managed services) Remote start is available.
Remote Managed Services will reduce the burden of monitoring, reporting, warning and management of IT operations that prevent switching to new technology. For more than 10 years, the mandate given to IT is to do more with less so that the staff was busy with the maintenance of the same technology. To transform the data center, IT staff must find time for training, planning and execution.
A group of IT experts who operate outside the Service Operations Center uses remote management devices can increase this capacity to the various installations at a cost of more acceptable and encourage faster return on investment.
No comments:
Post a Comment