SAN INFRASTRUCTURE FOR HIGH CAPACITY, PERFORMANCES, REDUNDANCE AND DATA PROTECTION
THE HGST AND DATACORE SOLUTION
Performance degrades during high work time, some SQL Server applications are not able to maintain or deliver the necessary, very high generalized latencies on the LUNs and the impossibility to maintain an appropriate QoS for the strategic applications.
Last but not least,there are infrastructural complexity and no disaster recovery site.
To create an appropriate infrastructure to support the productive and business continuity
To improve SAN performance for the strategic applications (typical SQL apps) and for virtual servers;
To maintain over time scalability needs in terms of space and performance and the integration with new technologies possible only by release form brand legacy hardware;
To migrate the virtualization platform in use to the new infrastructure without downtime;
To realize the business continuity solution by sharing the infrastructure specularly on two distinct data center synchronized in a synchronously manner
To prepare the option for the data asynchronous replay on the third site to implement a disaster recovery solution;
To improve current RPOs and RTOs related to backups by integrating SAN Continuous Data Protection capabilities for data recovery, including possible ramsonware infections
A NEW CONCEPT OF STORAGE: SAN VIRTUALIZATION
Software Defined Storage is the management and storage solution released from technology and brand links that are typical of traditional legacy storage environments.
In addition to becoming a single point of management and provisioning of all the hardware storage provided, the SDS typically integrates SAN virtualization business continuity data protecntion, advanced caching mechanisms (for example,autotiering on multiple levels) to increase considerably the performance of all or part of the storage, passing through advanced QoS and monitoring/alerting systems that make data management datacenters,even the most complex, easier and safer.
The SAN virtualization systems SDS, allows the aggregation of multiple disk portions (LUN), which can be supported on different brands and technologies in order to create a single data space (pool) on which to create Virtual Disks of greater or lesser capacity of that of the single initial LUNs. The software mirroring mechanisms present, allow the movement of Virtual Disk or the migration of data from legacy storage to the SDS environment, then from one hardware to another "hot" without interrupting production. The addition, removal of disk space or cache mechanisms is adaptable to different virtual disks or pools according to the dynamism and the requests of the business. The possibility of using caches based on RAM, SSD or of the latest generation make the SDS solution extremely performing that in some configurations can reach the order of millions of IOPs and several thousand MBps.
The Serv24 storage server platforms with JBOD Ultrastar Data60 from HGST by Western Digital are developed specifically for SDS solutions and the integration with the SANsymphony DataCore Defined Storage Software achieve them to reach their goals in order of performance, reliability, redundancy and data protection. The current nodes with VMware server virtualization are maintained, but existing storage is replaced. The HGST Serv24 system, is a Storage Server capable of containing up to 24 NVMe SS200 flash units, each with a capacity of up to 7.68TB of high endurance enterprise class. Thanks to the high flexibility it can be used either as a stand-alone Storage Server or flash storage or in combination with a scalable storage system.
Utilizing the SANsymphony SDS solution for storage virtualization, the full potential of the SERV24 unit, which is connected via SAS 12G to the JBOD HGST Ultrastar Data60 (can be equipped with up to 60 12TB capacity hard disks), provides a high capacity solution with extreme performances 720TB RAW of disk space and 184TB RAW on flash NVMe(the latter can be used as data space or high performance caching with the availability of millions of IOps on the entire infrastructure).
INFRASTRUCTURE SCHEME WITH DATACORE AND CACHING SSD
Integration with the SANsymphony solution simplifies management: automating several operations in a single management console, and by managing increasing amounts of data with an unchanged number of resources. The migration from the old legacy storage is performed using the mirroring system and the hot multipath system without interrupting the service. Business Continuity objectives are achieved by using the mirror mechanisms of the vDisk applied remotely (up to 100Km), the data center is divided into two mirrored units and redundantly activated / active (active servers on both sites. The SRV2U24 system provide a high performance buffer necessary for the creation of the third site for disaster recovery on which the asynchronous replication of all data is carried out in continuous mode. The advanced caching mechanisms on RAM and AutoTiering on Flash guarantee extremely high performance for VMs and critical applications, fully exploiting the potential of HGST SERV24 equipped with NVMeSN200 that reduce the execution times of some procedures to a tenth. The supplied QoS filters protect the rest of the VMs from potential saturations of the available band in the connection, ensuring adequate performance for each server and application.
Datacore Continuous Data Protection system allows very short RPOs revovery files, VMs Datastores or vDisks, with retention up to 14 days thanks to the NVMe SN200 unit on the HGST SERV24 server on which it is "supported"with no impacts on the general performances of the infrastructure. A third benefit from a combined solution of this type is the total cost reduction (TCO). The high reliability of HGST servers and JBOD, the high endurance of the flash units, combined with the extreme flexibility of the DataCore SANsymphony solution (the software is transferable on different hardware in case of obsolescence), leads to a fault reduction, by the modularity of the solution to a more dynamic and long-lasting use, and to an overall optimization of the resources used.
Storage Server HGST Serv24
Developed for Software Defined Storage
Server 2U Rack
Redundat power units 1200W 80+ Platinum
Double processor between: Intel Xeon Scalable 5120 (2,2Ghz, 14Core), Xeon Scalable 6140 (2,3Ghz 18Core), Xeon Scalable 8160 (2,1Ghz 24Core)
256GB RAM DDR4 2666Mhz
N. 24 slot NVMe 2,5”, possibility to use flash NVMe
2 PCIe Gen3 x16 and 1 PCIe Gen3 x8 slot
Connectivity JBOD with HBA/Controller RAID SAS 12G
Connettivity choice among LAN 10Gbit, 40Gbit, FC 16Gbit/32Gbit
HGST SN200 NVMe 3.2TB, performance for one single unit:
Sequential Read 3.350 MBps
Sequential Write 2.100 MBps
Random Read (max IOPS, 4KiB) 835.000IOps
Random Write (max IOPS, 4KiB) 200.000 IOps
Mixed Random Read/Write (max IOPS 70%R/30%W, 4KiB) 550.000IOps
JBOD HGST Ultra Data60
Developed for Software Defined Storage
JBOD 4U rack
Redundand power unit 1600W 80+ Platinum
N. 60 slot 3,5”/2,5” SAS3 12G for Hard Disks e SSDs
SAS3 12G module redundant Hot Swap, 6 Mini-SAS3 HD connection per module
Please contact us for more information