|
Hi all, I am working on an application that involves multiple processes
accessing and updating different databases: tables are split into 3 different
databases that are accessed by 3 different processes. This architectural
decision was made in the past because it has been noticed that former revisions
of postgres used to lock the whole database when a process performs an update
(on a record) which prevents the other processes from accessing the database. After some readings on postgres concurrent client management, I am
planning on changing the database structure to make all application processes
access a shared database (i.e. put all application tables in a single database
that would be shared among the 3 processes). I did not find in the
documentation (at least in the current versions of postgres starting from
revision 7) any specific problem related to doing this, comments on this
problem will be appreciated. Thanks! Hanan |
