PTKC

Overview

PTKC is a database instance level operation and maintenance tool that PTK generates separately after deploying a database cluster. PTK generates a copy of PTKC under the operating system user running the database on each server.

PTKC does not rely on any metadata , rely entirely on the environment variables of operating environment and database static configuration files and other information to identify the cluster , so you can copy directly to any cluster instance to use , but also through the PTK independent generation :

ptk gen-ptkc

The tool is supported since PTK V0.8.

Supported Subcommands

start: Start the cluster or instance
stop: Stop the cluster or instance
restart: Restart the cluster or instance
build: Rebuild a specified instance
status: Show status of db instance or cluster
show-guc: Print database parameters
show-hba: Print pg_hba.conf content of database
failover: Promote a standby server to primary.
switchover: Promote a standby server to primary and demote primary server to standby

Comparison Among ptkc, gs_ctl, and gs_om

Function	ptkc	gs_ctl	gs_om
Query the database status	Support for cluster and instance	Support for the local instance only	Support for cluster and instance
Start a database	Support for cluster and instance	Support for the local instance only	Support for cluster and instance
Stop a database	Support for cluster and instance	Support for the local instance only	Support for cluster and instance
Restart a database	Support for cluster and instance	Support for the local instance only	Support for cluster and instance
Rebuild a database	Supported on the local and remote servers	Support for the local instance only	Supported on the local and remote servers
Specify an upstream node to rebuild a database	Supported	Supported	Not supported
Failover	Supported on the local and remote servers	Support for the local instance only	Not supported
Switchover	Supported on the local and remote servers	Support for the local instance only	Not supported
Query database parameters	Supported on the local and remote servers	Not supported	Not supported
View HBA information	Supported on the local and remote servers	Not supported	Not supported

ptkc Usage

PTKC usage is kept consistent with PTK usage. PTK is a multi-cluster management tool and PTKC is a single-cluster O&M tool. Therefore, the PTKC instructions are more simple than those of PTK.

Note: PTKC itself does not store any metadata and completely depends on the existing status file to parse cluster information. Because the current status information does not include the system user information, PTKC does not support such scenarios as multiple nodes, multiple users, and multiple instances on a single server.

The following shows some examples. For more commands, run ptkc -h.

Query the Cluster Status
```
ptkc status
```
Restart a Cluster
```
ptkc restart
```
Perform Primary/Standby Switchover
```
ptkc switchover -H <IP>
```
Query Database Parameters
```
ptkc show-guc -k port
```
Show pg_hba.conf content
```
ptkc show-hba
```