When facing login failures with phpPgAdmin, verify these critical configurations:

# pg_hba.conf excerpt (usually in /var/lib/pgsql/9.1/data/)
# TYPE  DATABASE    USER        ADDRESS          METHOD
host    all         all         127.0.0.1/32     md5
local   all         postgres                      md5

After modifying pg_hba.conf, remember to reload PostgreSQL:

service postgresql-9.1 reload

Create a dedicated user with proper privileges instead of using postgres directly:

CREATE USER webuser WITH PASSWORD 'SecurePass123';
GRANT ALL PRIVILEGES ON DATABASE postgres TO webuser;
ALTER USER webuser CREATEDB;

The config.inc.php needs these essential settings:

$conf['servers'][0]['host'] = 'localhost';
$conf['servers'][0]['port'] = 5432;
$conf['servers'][0]['sslmode'] = 'allow';
$conf['servers'][0]['extra_login_security'] = false;

// For debugging purposes
$conf['debug'] = true;

If you still experience login failures:

1. Verify PostgreSQL is listening on the correct interface:

   netstat -tulnp | grep 5432

2. Test direct psql connection:

   psql -h 127.0.0.1 -U webuser -d postgres

3. Check PHP error logs:

   tail -f /var/log/httpd/error_log

• SELinux blocking connections (temporarily test with setenforce 0)
• PHP's pgsql extension not loaded (php -m | grep pgsql)
• Firewall rules blocking port 5432 or 80

For testing purposes, try connecting via UNIX socket:

$conf['servers'][0]['host'] = '';
$conf['servers'][0]['pg_dsn'] = 'pgsql:dbname=template1';

---

When phpPgAdmin stubbornly rejects valid credentials that work perfectly in psql, we're typically dealing with one of three fundamental issues:

// Common failure points:
1. pg_hba.conf misconfiguration
2. PHP/Apache environment issues
3. phpPgAdmin security restrictions

Your current MD5 setting is correct, but let's verify the complete authentication chain. The critical lines should resemble:

# TYPE  DATABASE        USER            ADDRESS                 METHOD
local   all             postgres                                peer
host    all             all             127.0.0.1/32            md5
host    all             all             ::1/128                 md5

After modifying pg_hba.conf, remember to reload PostgreSQL:

sudo service postgresql-9.1 reload

Beyond the basic settings you've already adjusted, these additional configurations often resolve login issues:

$conf['servers'][0]['port'] = 5432;
$conf['servers'][0]['sslmode'] = 'allow';
$conf['servers'][0]['pg_dump_path'] = '/usr/bin/pg_dump';
$conf['servers'][0]['pg_dumpall_path'] = '/usr/bin/pg_dumpall';

The webuser needs explicit login privileges. Verify with:

psql -U postgres -c "SELECT rolname, rolcanlogin FROM pg_roles WHERE rolname = 'webuser';"

If missing, grant access:

psql -U postgres -c "ALTER ROLE webuser WITH LOGIN;"

CentOS 6.2's default PHP settings might interfere with session handling. Check /etc/php.ini:

session.save_path = "/var/lib/php/session"
session.save_handler = files

Ensure the session directory exists and has proper permissions:

sudo mkdir -p /var/lib/php/session
sudo chown -R apache:apache /var/lib/php/session

Verify PHP can actually reach PostgreSQL:

Check these critical logs when issues persist:

tail -n 50 /var/log/httpd/error_log
tail -n 50 /var/lib/pgsql/9.1/data/pg_log/postgresql-*.log