During MWC 2017 ZTE will show some of its best IT hardware products. Among this portfolio, we will see the 2-channel rack server R5300G4, 4-channel rack server R8500G4, blade server E9000, and unified storage KU5200. To effectively provide hardware support for NFV, data centres, cloud storage and other solutions, all these devices are featured in 100G switching, intelligent acceleration, liquid cooling, HVDC, modularization, high-capacity storage and so on. The machines displayed use the latest generation Intel Purley platform, announced six months earlier than rival vendors.
R5300G4 is ZTE’s new generation 2U 2-socket rack server, supporting Intel’s latest Purley platform processor.
The R5300G4 is available in 8 disk-position, 12 disk-position and 24 disk-position formats, and the controller uses a modular design. Meaning the user can replace the controller as needed to reduce the expansion cost.
To meet the growing demand for massive-capacity storage, R5300G4 maximally supports 30 hard drives, 24 disk-position controller at frontend, and 6 hard drives at the backend. One machine can support up to 240T.
R5300G4 is I/O oriented with a large choice of interface: 9* PCIe slots, 6* USB, 2* VGA. Unlike other vendors, the R5300G4 VGA interface is on the frontend, which is making it more convenient to maintain.
To match to the low-energy data centre construction requirement, the R5300G4 supports liquid cooling and HVDC; according to the machine’s own energy consumption, it dynamically adjusts fan speed to save energy.
ZTE’s new generation 4U 4-socket rack server is supporting latest Intel’s Purley platform series processors. The R8500G4 is available in 16 disk-position and 28 disk-position formats. The new modular design of the controller allows the user to replace the controller unit as needed.
To follow the increasing demand for high-capacity storage, the R8500G4 maximally supports 28 hard drives (including 8 NVME disks). This is an industry record.
ZTE’s self-developed blade server represents the best technical mix of compute, storage and network capabilities. It’s supporting 2-socket half-height blade and 4-socket full-height blade. It’s keeping up with Intel’s release cycle to provide the latest platform server products, and having won the highest industrial design award in China – the Red Star Award.
Faced with the increasing data processing needs of cloud computing and big data technologies, E9000 provides a 100G bandwidth solution that drastically improves the network performance of the data centre. The E9000 includes an intelligent acceleration module, which can effectively match up the acceleration of a virtualization switch module.
To reach the low-energy data centre construction mandate, E9000 supports liquid cooling, HVDC and power capping; according to the machine’s own energy consumption, it dynamically adjusts the fan speed, to save energy.
ZTE’s self-developed high-end unified storage product, supporting 48* 1Gb iSCSI, 24* 10Gb iSCSI, 48* 8Gb FC, 24* 10Gb FCoE, 24* 24Gb miniSAS, and 24* 48Gb mini-HD SAS, super interface capability, 1T cache, 1728 hard drives and super scalability to meet every scenario needs.
It supports ZTE EEP technology, delivering higher reliability and faster RAID reconfiguration with a higher concurrent access performance than traditional RAID.
It supports the Cinder interface, achieving seamless interoperation with Openstack.
It provides disk roaming. During customer maintenance, to improve maintenance efficiency, users can plug the hard drive with no need to record the specific location of the disk.
ZTE’s distributed storage system ZXCLOUD KS10000 is based on the industry-leading distributed and auto-expandable cloud storage Ceph.
It supports unified storage, providing external object storage, block storage, and file storage services. It also provides flexible expansion modes, including independent expansion of hard disk and storage nodes, or simultaneous expansion.
KS10000 supports multi-cluster management, no single point of failure impact within the system. When one node or one hard drive fails, it will automatically be isolated from the cluster, not affecting the use of the whole system service; it supports multi-copy backup/erasure code mechanism to ensure data reliability; load balancing, controllable impact of data migration.