Re: Informix vs. Sybase vs. Oracle vs. (gasp) MS SQL Server

David Williams wrote: > > In article <3481D92B.75AA9F0D@voicenet.com>, "Scott C. Gray" > <ickyspam_gray@voicenet.com> writes > >> page level locks give HIGHER PERFORMANCE yes BItUT the reduce > >> concurrency and hence LOWER RELIABILITY. > > > >Be careful with your wording...locking granularity has nothing at > >all to do with reliability (indeed, table locking vs. column locking > >has the same reliability when properly implemented). It is all > >simply a matter of concurrency and efficiency. > > Most applications produce an error when a lock error occurs. > I assume that you are referring to deadlocks. While it is true that deadlocks may be reduced in a row vs. page locking environment, they can still occur (don't forget we are only talking about locking... not versioning--which can also be done in a page locked environment). It is foolish for any application developer to fail to check for such conditions (although, in my experience, almost everyone does). Thus, for a properly coded application, page locking does not reduce application reliability one bit...it just increases the likely hood that the proper error detection/recovery code will be executed (although, in a properly developed environment, it is possible to eliminate deadlocks anyway). > 1. Queries fails > 2. Database modifications produce an error and the transaction is > rolled back (few applications retry). > > Hence APPLICATION RELIABLITIY is reduce. Once again...be careful about your wording. It seems that you are producing a false analogy along the lines of: 1. Some conditions cause the database to generate an error. 2. Many applications don't check for these errors The database is not working properly. Do not equate sloppy application development with deficiencies in the tools being used to develop them. -scott -- Scott C. Gray gray@voicenet.com "my keybard is brken" Sybase Professional Services scott.gray@sybase.com http://www.voicenet.com/~gray/sqsh.html