The next four months are crucial for Microsoft as it
puts the finishing touches to the much anticipated update of its
SQL Server relational database, due in November.
Microsoft is telling users the five-year wait since SQL Server
2000 began shipping will be worthwhile. And delegates at Tech Ed
Europe were told that Microsoft’s strategy for SQL Server 2005
would be to address performance concerns that have limited the use
of its predecessor in the datacentre, where Oracle and DB/2 are the
preferred choice of databases that power mission-critical business
systems.
Mauro Meanti, Microsoft general manager for the EMEA server
division at Microsoft said, “We are focusing on the high end with
SQL Server to scale to the top 5% of enterprise [users].” Such
organisations have traditionally turned to Unix databases in order
to meet the demands of running global operations.
Euan Garden, group program manager for SQL Server, said users
can expect a boost in performance when multi-core chips like
Intel’s 64-bit multi-core Monticeto chip begin shipping. “With
Monticeto and x64 chips, what you will get out of two-way and
four-way [servers] will be tremendous,” he said.
Garden also discussed some of the SQL Server management
improvements Microsoft has made to lower the cost of database
administration. These include database mirroring for automatic
failover if the database server crashes; a snapshot database
restore facility and the ability to install extra memory on a
server running SQL Server 2005 on Windows Server 2003 without
requiring a reboot.
Security, as in other Microsoft server products, will now be a
priority with high-risk database functions switched off by default.
Microsoft is also undertaking certification of SQL Server 2005
under the US NSA EAL4+ programme, a high-level of security
accreditation required by governments.
Another performance improvement is the ability to write code
using Visual Basic and C# within the database in order to perform
computationally intensive tasks. Such tasks cannot usually be
achieved efficiently using SQL.