Sql:sybase and Oracle Comparison

1. Sybase and Oracle Comparison http://blog.itpub.net/14067/viewspace-1030014/

Oracle employs a multi-threaded, multi-process architecture

Sybase uses a single-process multi-thread architecture

Both Oracle and Sybase use multiple threads.

The multi-clue mode can manage a large number of user processes with fewer clues, and the thread process is dynamically adjustable, and when the number of users increases, the thread process automatically increases, and the thread process automatically decreases when the number of users decreases. The multi-clue structure greatly reduces the utilization of system resources by Oracle and Sybase, and improves the usage of system resources.

For Oracle's architecture, in a centralized multiuser environment, each user consumes a single server process, and in Client/server and Internet mode, multiple user sharing processes can be used to execute user requests in a multi-threaded manner. In addition, for special-level users, you can use a single server process to achieve the best level of user requests. Oracle's server processes are used to process database requests at the same time, they take requests from the request queue, perform parsing, permission checks, and a range of internal operations, and coordinate Oracle's other background processes to work together to complete the processing of the request and return the results of the operation to the answer queue. Oracle Multi-server process structure, can realize the parallel processing of database transaction, improve the response speed of concurrent transaction processing. and the multi-server structure has the very flexible expansibility, when the hardware platform processing power increases, the number of server processes can also increase, the database performance also increases. Multi-server structure avoids the server process bottleneck which is easily caused by single server structure, and avoids the phenomenon of single server process deadlock caused by it.

Sybase uses a single-process multi-thread approach, and user requests are managed by a process for total scheduling and management. In the single-server and single-client/server mode, its management of the database is relatively simple, the memory resource occupies less, in the case of fewer concurrent users, the utilization of system resources is higher than the multi-process method, but at the same time, when the number of concurrent users reaches a certain number , the system processing performance is greatly reduced, and the server process bottleneck and deadlock, and the system can be very narrow in the scope of expansion. It is technically impossible to provide support for large multiprocessor systems with limited performance.

Although you can increase the number of dataserver processes by configuring the maximum number of engines, the maximum number of CPUs cannot be exceeded.

The principle of processing is

The CPU is the resource that the operating system can mobilize, and Sybase itself does not allow the database to work on which CPU,
Using the engine will run several processes,
If the engine is 1,sybase only 1 processes, the operating system will not be able to run it on two CPUs
If the engine is 2,sybase up to 2 processes,

Sybase decides which dataserver process to assign the task to, and the operating system can allow two processes to run on two CPUs.

If the OS is primarily running Sybase, start with 2 suitable

does not work at the user level, as Oracle does.

2.sybase vs. Oracle stored procedures http://blog.csdn.net/daniel112/article/details/2304288

There is some inconsistency between creating stored procedures in Oracle and the syntax under Sybase and SQL Server.

Sql:sybase and Oracle Comparison