We already have wait_on_release_in_recovery kind of parameter, namely recovery_timeout:
# maximum time in seconds to wait for the recovering node's postmaster # start-up. 0 means no wait. # this is also used as a timer waiting for clients disconnected before # starting 2nd stage recovery_timeout = 90 -- Tatsuo Ishii SRA OSS, Inc. Japan > > Ok, what about having new switch: > > > > client_idle_limit_in_recovery > > > > BTW, do we have to kill clients immediately even if they are *not* > > idle in online recovery as Tiago said? > > > [cut] > > I think that the following mechanism would be nice. > > It is easy to include into the first stage recovery script some commands > sending notifications to clients informing that connections shall be > relesed. Then we can wait gracefully some time (wait_on_release_in_recovery) > in second stage checkpoint that they are really relesed. If the grace period > passed and we have still open client connections, then next boolean variable > (release_clients_in_recovery) give us permission to shutdown immediately > remianed connections. If we are not permitted to do, then recovery process > can simply failed. > > Unfortuntely I have no idea what to do for long transactions (connection in > use in transaction) :( > > regards, > Edward _______________________________________________ Pgpool-general mailing list [email protected] http://pgfoundry.org/mailman/listinfo/pgpool-general
