[ https://issues.jboss.org/browse/JBJCA-1361?page=com.atlassian.jira.plugin... ] Stephen Fikes updated JBJCA-1361: --------------------------------- Description: http://www.ironjacamar.org/doc/userguide/1.2/en-US/html_single/index.html... It wasn't immediately clear to me how {{EntirePool}} and {{IdleConnections}} flush strategies work since both use the word "all" in their explanation/definition. Comparing {{AllConnections}} and {{AllIdleConnections}} it is clear that "all" with regard to {{EntirePool}} and {{IdleConnections}} is limited in some sense. The subtlety lies in the fact that a single datasource can have multiple (internal) sets of connections grouped/segregated by authentication credentials. Consequently, "all" in the context of {{EntirePool}} or {{IdleConnections}} is limited to a subset of the connections - specifically to those that are associated with a specific authentication credential whereas the {{All*}} variants include all subsets associated with the datasource (regardless of credentials). was: http://www.ironjacamar.org/doc/userguide/1.2/en-US/html_single/index.html... It wasn't immediately clear to me how {{EntirePool}} and {{IdleConnections}} flush strategies work since both use the word "all" in their explanation/definition. Comparing {{AllConnections}} and {{AllIdleConnections}} it is clear that "all" with regard to {{EntirePool}} and {{IdleConnections}} is limited in some sense. The subtlety lies in the fact that a single datasource can have multiple (internal) sets of connections grouped/segregated by authentication credentials. Consequently, "all" in the context of {{EntirePool}} or {{IdleConnections}} is limited to a subset of the connections that are associated with a specific authentication credential whereas the {{All*}} variants are not so limited. -- This message was sent by Atlassian JIRA (v7.5.0#75005)