What if we need to take only schema backup without data? (pg_dump -s)

What if we need to take a specific schema backup with database?(pg_dump -n

Hi, I am using postgres v14
And I have downloaded github.com/habx/pg-commands v0.6.1 in my golang backend project.
I have set the correct env path to postgresql bin directory.
But if I run the backend, I got pg_dump: illegal option -- blob error when I attempt to dump.
Please, help!

I encountered a race condition detected from go test and after reviewing your CircleCI Orb i see you have it turned off. It might be worth enabling it to catch those cases.

Orb Source line 312+

- go/test:
                covermode: atomic
                coverprofile: coverage.out
                failfast: true
                race: false
                verbose: true

Race condition was encountered between

Read at 0x00c0003fc020 by goroutine 67:
  github.com/habx/pg-commands.(*Dump).Exec()
      /x/x/go/pkg/mod/github.com/habx/[email protected]/pg_dump.go:64 +0xb16

and

Previous write at 0x00c0003fc020 by goroutine 68:
  github.com/habx/pg-commands.streamOutput()
      /x/x/go/pkg/mod/github.com/habx/[email protected]/utils.go:45 +0x1d3
  github.com/habx/pg-commands.(*Dump).Exec.func1()
      /x/x/go/pkg/mod/github.com/habx/[email protected]/pg_dump.go:54 +0x8a

I was running a unit test which uses suite.run to sequentially trigger multiple database exports. In Exec the streamOutput function is started as a go routine and can not be observed from outside to know when it has finished. I believe that is where the race condition occurs as it is unmanaged writing and reading from the Result.Error variable in both routines

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Pending Status Checks

These updates await pending status checks. To force their creation now, click the checkbox below.

• fix(deps): update module github.com/go-pg/pg/v10 to v10.11.0

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

As part of a system I'm working on we're dumping (and restoring) some fairly large databases (+50GB). Obviously that takes some time and we would like to keep the users aware of the progress.

The problem I'm encountering is that the output destination is hard coded to stdout as seen here.

Please let me know if there's a better way of solving the above issue. But I think this would be best solved by adding a io.Writer to the ExecOptions struct. It should obviously default to os.Stdout but it could be used to stream the output to multiple destinations ranging from a simple file to a database while the dump is taking place.

Would you guys be interested in a PR implementing the above feature?

I am attempting to build an all in one utility to facilitate in backup/restore of various remote databases to/from object stores such as S3/MinIO/Azure Blob and storing all transactions in Redis. We recently updated our PostgresQL versions to 16, and I cannot use pg_dump in this package as the version is 14

2024/03/03 16:40:54 error: &{Err:exit status 1 CmdOutput:pg_dump: error: server version: 16.2 (Debian 16.2-1.pgdg120+2); pg_dump version: 14.11 (Ubuntu 14.11-0ubuntu0.22.04.1)

Can I update the version used in this package?

As the title indicates, I get no output despite creating my object as below.

dumpExec := dump.Exec(pg.ExecOptions{StreamPrint: true})

I'm using version v0.3.0.

Please let me know if I should post the rest of my code or if there's anything else I can do to assist.

Hi, I have a problem. Console error this like following
panic: command not found: pg_dump

What is problem?

go version go1.19.5 windows/amd64

CODE

package main

import (
"fmt"
pg "github.com/habx/pg-commands"
"os"
)

func main() {

fmt.Println("backupp")

dump, err := pg.NewDump(&pg.Postgres{
	Host:     "localhost",
	Port:     5432,
	DB:       "dev_example",
	Username: "example",
	Password: "example",
})
if err != nil {
	panic(err)
}

dump.EnableVerbose()

dumpExec := dump.Exec(pg.ExecOptions{StreamPrint: true, StreamDestination: os.Stdout})
if dumpExec.Error != nil {
	fmt.Println(dumpExec.Error.Err)
	fmt.Println(dumpExec.Output)

} else {
	fmt.Println("Dump success")
	fmt.Println(dumpExec.Output)
}

restore, err := pg.NewRestore(&pg.Postgres{
	Host:     "example",
	Port:     5432,
	DB:       "dev_example_new",
	Username: "example",
	Password: "example",
})
if err != nil {
	panic(err)
}
restoreExec := restore.Exec(dumpExec.File, pg.ExecOptions{StreamPrint: false})
if restoreExec.Error != nil {
	fmt.Println(restoreExec.Error.Err)
	fmt.Println(restoreExec.Output)

} else {
	fmt.Println("Restore success")
	fmt.Println(restoreExec.Output)

}

}

I attempted to dump a few small tables. The actions occurred completed so fast that the timestamp used to ensure uniqueness were all the same. It would be great to expose a method that enables a user to specify the full filename

As in the title.

Running the following code, based on the example:

func backupPostGres(host, user, pass, db, backupDir string, port int) error {
	filename := "backup.tar"
	dump := pg.NewDump(&pg.Postgres{
		Host:     host,
		Port:     port,
		DB:       db,
		Username: user,
		Password: pass,
	})
	dump.SetPath(backupDir)
	dump.SetFileName(filename)
	res := dump.Exec(pg.ExecOptions{StreamPrint: false})
	if res.Error != nil {
		return res.Error.Err
	}
	fmt.Println(res.Output)
	return nil
}

Yielded no errors and no log output.

When the binary is present and returns an error, the above code simply returns the error message exit status 1. It would be worth including wrapping the error with Error.CmdOutput and logging this in the example. e.g:

	if res.Error != nil {
		err := fmt.Errorf("%w: %s", res.Error.Err, res.Error.CmdOutput)
		log.Print(err)
		return err
	}