We have 2 vFabric Postgres 9.2.4 virtual appliances running as a master slave cluster to provide a database for Horizon Workspace.
The horizon database size is reported at 34 MB On the master server the drives filled up (taking the database offline) at the following locations:
/dev/sdb 7.8 GB (/var/vmware/vpostgres/9.2/pgdata/archive is full of 16MB files)
/dev/sde 1.9 GB (/var/vmware/vpostgres/9.2-archive is full of 16 MB files)
show_replication_status script gives the following on the master server:
sync_priority | slave | sync_state | log_receive_position | log_replay_position | receive_delta | replay_delta
---------------+--------------------------+------------+----------------------+---------------------+---------------+--------------
0 | gr2cmnxvfpv02.cmn.net | async | 2/48696CB8 | 2/48696AB0 | 0 | 520 (1 row)
What is causing the 2 drives to fill up so quickly? And how can we resolve.
