Entware how to auto start pgsql

[XIYO]

Here's the English translation of your text:

Hello, I'm trying to run pgsql on my AC88U(386.11) router.
I know it may not be the best idea, but I want to give it a try.

## I've encountered several issues but have managed to resolve them all.

- Search for `pgsql` in opkg, not `postgresql`.
- `pg_ctl` cannot be run as the root user, so you have to use the `-U` option to change the user.
- Accounts get reset, so I had to create scripts to register accounts (passwd.add, group.add... etc.)
- `pg_ctl` logs issues upon running, so where it is run from is also important.

I've resolved various issues, including allowing external access, and have successfully launched the service.

## Now, the final challenge remains.

### The first method successfully launches the service.

# In a directory where 'postgres' has write permission
pg_ctl -U postgres -D /mnt/sda1/entware-data/var/pgsql start

### The second method also successfully launches the service.
I've also modified S98postgresql, and it runs well (although the logs aren't pretty...).

S98postgresql

#!/bin/sh

# Installation prefix

prefix="/mnt/sda1/entware-data"

# Data directory
PGDATA="$prefix/var/pgsql"

PG_DUMP_DIR="$prefix/var/pgsql/15.2/dump"

# Who to run the postmaster as, usually "postgres".  (NOT "root")
PGUSER=postgres

# Where to keep a log file
PGLOG="$PGDATA/serverlog"

# What to use to shut down the postmaster
PGCTL="$prefix/bin/pg_ctl"

# What to use to start up the postmaster (we do NOT use pg_ctl for this,
# as it adds no value and can cause the postmaster to misrecognize a stale
# lock file)

DAEMON="/opt/bin/pg_ctl"

#########
if [ -z "$1" ] ; then
    case `echo "$0" | /bin/sed 's:^.*/\(.*\):\1:g'` in
        S??*) rc="start" ;;
        K??*) rc="stop" ;;
        *) rc="usage" ;;
    esac
else
    rc="$1"
fi

case "$rc" in
    start)
        echo -n "Starting PostgreSQL: "
        cd $PGDATA
        $DAEMON -U $PGUSER -D $PGDATA -l $PGLOG start
        echo ok
        ;;
    stop)
        echo -n "Stopping PostgreSQL: "
        $DAEMON -U $PGUSER -D $PGDATA stop
        echo ok
        ;;
    restart)
        echo -n "Restart POstgreSQL: "
        $DAEMON -U $PGUSER -D $PGDATA restart
        echo ok
        ;;
    dumpall)
        PG_DUMP_FILE=$PG_DUMP_DIR/dumpall-`date +%FT%H%M%S`
        if [ -x $prefix/bin/gzip ]; then
            PG_DUMP_FILE=$PG_DUMP_FILE.gz
            echo -n "PostgreSQL pg_dumpall and gzip to $PG_DUMP_FILE "
            su - $PGUSER -c "$prefix/bin/pg_dumpall | $prefix/bin/gzip -c > $PG_DUMP_FILE"
        else
            echo -n "PostgreSQL pg_dumpall to $PG_DUMP_FILE "
            su - $PGUSER -c "$prefix/bin/pg_dumpall > $PG_DUMP_FILE"
        fi
        echo ok
        ;;
    *)
        echo "Usage: $0 (start|stop|restart|dumpall|usage)"
        ;;
esac

/opt/etc/init.d/S98postgresql start

This command works fine.

My final challenge is to ensure that the service starts automatically after a reboot.

In /jffs/script/services-start,

#!/bin/sh
/jffs/addons/amtm/shellhistory # Added by amtm
/opt/etc/init.d/rc.unslung start # Added by xiyo

I've written it like this, but while the nginx I installed for comparison starts normally, PostgreSQL does not. I suspect it's a permissions issue, but I don't even know how to check the logs...

Any help would be greatly appreciated.

I wish everyone a happy day.

 

[Martinski]

...

/opt/etc/init.d/S98postgresql start

This command works fine.

My final challenge is to ensure that the service starts automatically after a reboot.

In /jffs/script/services-start,

#!/bin/sh
/jffs/addons/amtm/shellhistory # Added by amtm
/opt/etc/init.d/rc.unslung start # Added by xiyo

When the "/jffs/scripts/services-start" script runs during a reboot, there's no guarantee that the USB-attached disk drive partition (where Entware is installed) has been fully mounted every time, so the recommended place to start Entware services is the "/jffs/scripts/post-mount" script.

If you installed Entware via the built-in AMTM utility, the "/jffs/scripts/post-mount" script should already exist and include the "/jffs/addons/diversion/mount-entware.div" script which mounts Entware and starts all service scripts found in the "/opt/etc/init.d" directory. If, for some reason, this is not working for you, first make sure that the code is included in your "post-mount" script:

. /jffs/addons/diversion/mount-entware.div

If it's not found or not working for your specific case, the "post-mount" script is still the place where you should add your own code to start the specific Entware service that you want to run and troubleshoot.

HTH

 

[XIYO]

Thank you. It was resolved all at once.


What I did was:
- Made slight modifications to `/opt/etc/init.d/S98postgresql`

# add post-mount-start sciprt
    post-mount-start)
        echo -n "Starting PostgreSQL: "
        cd $PGDATA
        $DAEMON -U $PGUSER -D $PGDATA -l $PGLOG start
        echo ok
        ;;

# change start script
    start)
        echo "This script should not be run before the storage is mounted."
        echo "Run manually with '/opt/etc/init.d/S98postgres post-mount-start',"
        echo "or add the script to /jfft/script/post-mount."
        ;;

- Added the execution command to `/jfft/script/post-mount`

# added last line
/opt/etc/init.d/S98postgresql post-mount-start