- Integrated Systems
- Client Computing Devices
- Fujitsu Produkt-Supportservices
- Fujitsu Produkt-Supportservices
Mainichi communications: May 13, 2010*
Fujitsu is holding its annual event "Fujitsu Forum 2010" at Tokyo International Forum from May 13th to May 14th.
The schedule includes 111 seminars and around 150 different demonstrations. One of the planned presentation themes is something The MYCOM Journal Enterprise channel has been focusing on; cloud computing supported
by Fujitsu's storage solutions.
In this article, we will introduce a database integration solution using ETERNUS DX series SAN disk storage systems, which was exhibited in the demonstration hall.
The database integration solution using ETERNUS DX series, achieves a high-speed integrated database with low-power consumption. The demonstration displays to what extent a combination of SPARC Enterprise M3000 and ETERNUS DX80 systems can improve database performance.
In recent years, the construction of private cloud environments and integration of enterprise systems have been attracting attention as ways to reduce the TCO of enterprise ICT infrastructure. But in actually reducing TCO this way requires configuring of a large-scale and high-performance integrated database.
Fujitsu's database integration solution achieves TCO reduction in a much simpler way. By using SSDs as memory devices in the disk storage systems, the solution enables high-speed database access and significant power reductions. This is due to SSDs having outstanding excellence in random access performance and many advantages in terms of power consumption, heat generation, quake/impact tolerance and environmental robustness over HDDs (Hard Disk Drives).
Fujitsu's ETERNUS DX series is a suite of high-reliability and high-performance SAN disk storage systems that support both HDDs and SSDs as memory devices. This means by using ETERNUS DX series for your integrated database component, high-speed database access can be achieved by taking advantage of the attributes of SSDs.
The system uses HDDs for storage data file area as with traditional storage systems. But it can also use SSDs as Flash Cache area for use as server memory expansion. In traditional systems, frequently accessed data (hot data) is stored in server memory and infrequently accessed data on HDDs. This requires large-capacity memory or additional HDDs to improve data access performance. But using ETERNUS DX series, even with small servers, can minimize HDD access and achieve higher performance by use of SSDs for server memory expansion.
The new "Database Smart Flash Cache" function of database server "Oracle Database 11g Release 2" enables memory expansion using SSD. The Database Smart Flash Cache function eliminates disk I/O bottlenecks by moving hot data to flash memory transparently. With this function and SSDs mounted in ETERNUS DX series, database performance can be easily improved without changes to database settings.
Database Smart Flash Cache is now supported by Solaris OS and Oracle Enterprise Linux. SPARC Enterprise servers which support Solaris OS + ETERNUS DX series can therefore operate the entire database system. The verification result of this combination, as performed by Fujitsu and Oracle, is provided in a whitepaper.
At the Fujitsu demonstration booth, throughput and response time improvement was demonstrated using an actual system with the following configuration.
ETERNUS DX series is available in three ranges: entry-level ETERNUS DX60/DX80/DX90, midrange-level ETERNUS DX400 series, and enterprise ETERNUS DX8000 series. With SSDs supported on all except DX60, the above configuration can be implemented regardless of system size. Plus, Flash Cache can be added to existing databases just by using SSDs. In addition to hardware, Fujitsu's dedicated Oracle database team provides backup support in cooperation with Oracle. It also offers consulting services to support solution introduction and accompanying database migration and tuning.
*Translation of an article printed in Mainichi Communications Inc's MYCOM journal, published on May 13th, 2010. The content of this article was correct at the time of publication.
Diese Seite teilen