Manage the PXF configuration and the PXF Service instance on all Greenplum Database hosts.
pxf cluster <command> [<option>]
where <command>
is:
help
init (deprecated)
migrate
prepare
register
reset (deprecated)
restart
start
status
stop
sync
The pxf cluster
utility command manages PXF on the coordinator host, standby coordinator host, and on all Greenplum Database segment hosts. You can use the utility to:
$PXF_BASE
runtime configuration directory.$PXF_CONF
configuration to $PXF_BASE
.pxf cluster
requires a running Greenplum Database cluster. You must run the utility on the Greenplum Database coordinator host.
If you want to manage the PXF Service instance on a specific segment host, use the pxf
utility. See pxf
.
pxf cluster
help message and then exit.
register
command.
$PXF_CONF
directory to
$PXF_BASE
on each Greenplum Database host. When you run the command, you must identify the PXF 5 configuration directory via an environment variable named
PXF_CONF
. PXF migrates the version 5 configuration to
$PXF_BASE
, copying and merging files and directories as necessary.
pxf-log4j.properties
customizations to the
pxf-log4j2.xml
file.
$PXF_BASE
directory on each Greenplum Database host. When you run the command, you must identify the new PXF runtime configuration directory via an environment variable named
PXF_BASE
. PXF copies runtime configuration file templates and directories to this
$PXF_BASE
.
$GPHOME
be set, and is run once after you install PXF 6.x the first time, or run after you upgrade your Greenplum Database installation.
$PXF_BASE
) from the coordinator host to the standby coordinator host and to all Greenplum Database segment hosts. By default, this command updates files on and copies files to the remote. You can instruct PXF to also delete files during the synchronization; see Options below.
The pxf cluster sync
command takes the following option:
Stop the PXF Service instance on the coordinator host, standby coordinator host, and all segment hosts:
$ pxf cluster stop
Synchronize the PXF configuration to the standby coordinator host and all segment hosts, deleting files that do not exist on the coordinator host:
$ pxf cluster sync --delete