add-node workflow: fix recovery and wal restore issues, improve clean up logic

cli/scripts/install.py

@@ -3,7 +3,7 @@
 
 import sys, os, tarfile, platform
 
-VER = "25.0.0-alpha3"
+VER = "25.0.0-alpha4"
 REPO = os.getenv("REPO", "https://pgedge-upstream.s3.amazonaws.com/REPO")
 
 if sys.version_info < (3, 9):

cli/scripts/util.py

@@ -4,7 +4,7 @@
 import os
 import time
 
-MY_VERSION = "25.0.0-alpha3"
+MY_VERSION = "25.0.0-alpha4"
 MY_CODENAME = ""
 
 DEFAULT_PG = "16"

devel/setup/Dockerfile.rocky810

@@ -7,8 +7,9 @@ USER root
 
 ENV install="dnf install -y --allowerasing"
 RUN $install dnf-plugins-core
+RUN $install epel-release
 RUN $install python3 python3-pip git wget curl pigz which zip sqlite
-RUN $install openssh-server systemd sudo
+RUN $install openssh-server systemd sudo inotify-tools lsof
 
 RUN useradd build  -U -m -d /home/build  \
     && echo "build  ALL=(ALL)  NOPASSWD: ALL" >> /etc/sudoers

devel/setup/Dockerfile.rocky95

@@ -7,8 +7,9 @@ USER root
 
 ENV install="dnf install -y --allowerasing"
 RUN $install dnf-plugins-core
+RUN $install epel-release
 RUN $install python3 python3-pip git wget curl pigz which zip sqlite
-RUN $install openssh-server systemd sudo
+RUN $install openssh-server systemd sudo inotify-tools lsof
 
 RUN useradd build  -U -m -d /home/build  \
     && echo "build  ALL=(ALL)  NOPASSWD: ALL" >> /etc/sudoers

devel/setup/compose/Dockerfile.node.rocky95

@@ -31,7 +31,7 @@ RUN chmod 700 /home/pgedge/.ssh/
 RUN sudo chown pgedge:pgedge /home/pgedge/.ssh/id_rsa
 RUN sudo chmod 600 /home/pgedge/.ssh/id_rsa
 
-RUN sed -i '1iexport PGBACKREST_REPO1_CIPHER_PASS=Really_s3cure_password' ~/.bashrc
+RUN sed -i '1iexport PGBACKREST_REPO1_CIPHER_PASS=supersecret' ~/.bashrc
 
 USER root
 EXPOSE 22

devel/setup/compose/Dockerfile.node.ubuntu2204

@@ -31,7 +31,7 @@ RUN chmod 700 /home/pgedge/.ssh/
 RUN sudo chown pgedge:pgedge /home/pgedge/.ssh/id_rsa
 RUN sudo chmod 600 /home/pgedge/.ssh/id_rsa
 
-RUN echo 'export PGBACKREST_REPO1_CIPHER_PASS=Really_s3cure_password' >> ~/.bashrc
+RUN echo 'export PGBACKREST_REPO1_CIPHER_PASS=supersecret' >> ~/.bashrc
 
 USER root
 EXPOSE 22

env.sh

@@ -1,5 +1,5 @@
-hubV=25.0.0-alpha3
-hubVV=25.0.0-alpha3
+hubV=25.0.0-alpha4
+hubVV=25.0.0-alpha4
 
 aceV=$hubV
 kirkV=$hubV

src/backrest/backrest.py

@@ -7,8 +7,6 @@
 import sys
 from datetime import datetime
 from tabulate import tabulate
-import yaml
-from typing import Optional
 
 def pgV():
     """Return the first found PostgreSQL version (v14 thru v17)."""
@@ -94,7 +92,7 @@ def create_stanza(stanza, verbose=True):
 
     # Modify postgresql.conf to ensure archiving is on
     modify_postgresql_conf(
-        stanza, config['pg1-path'], config['repo1-path'], config['repo1-type']
+        stanza, config['pg1-path'], config['repo1-path'], config['repo1-type'], config['repo1-cipher-type']
     )
     # Modify pg_hba.conf for replication, if needed
     modify_hba_conf()
@@ -183,6 +181,7 @@ def backup(stanza, type="full", verbose=True):
         )
     else:
         util.message(f"Successfully completed {type} backup for stanza '{stanza}'")
+
 def restore(stanza, data_dir=None, backup_label=None, recovery_target_time=None, verbose=True):
     """Restore a database cluster to a specified state."""
     config = fetch_config()
@@ -254,6 +253,19 @@ def pitr(stanza, data_dir=None, recovery_target_time=None, verbose=True):
     if restore(stanza, data_dir, None, recovery_target_time, verbose):
         _configure_pitr(stanza, data_dir, recovery_target_time)
 
+def cleanup_replica(pg1_path):
+    """Cleanup the replica configuration and restore remnants."""
+    conf_file = os.path.join(pg1_path, "postgresql.conf")
+    changes = {
+        "hot_standby": "off",
+        "primary_conninfo": "",
+        "archive_command": "",
+        "archive_mode": "off"
+    }
+    for key, value in changes.items():
+        change_pgconf_keyval(conf_file, key, value)
+
+
 def _configure_pitr(stanza, pg_data_dir=None, recovery_target_time=None):
     """Configure PostgreSQL for point-in-time recovery."""
     config = fetch_config()
@@ -264,7 +276,6 @@ def _configure_pitr(stanza, pg_data_dir=None, recovery_target_time=None):
     config_file = os.path.join(pg_data_dir, "postgresql.conf")
     changes = {
         "port": "5433",
-        "log_directory": os.path.join(pg_data_dir, "log"),
         "archive_command": "",
         "archive_mode": "off",
         "hot_standby": "on",
@@ -288,6 +299,7 @@ def change_pgconf_keyval(config_path, key, value):
                 file.write(line)
         if not key_found:
             file.write(f"{key} = '{value}'\n")
+
 def create_replica(stanza, data_dir=None, backup_label=None, verbose=True):
     """Create a replica by restoring from a backup."""
     if restore(stanza, data_dir, backup_label, verbose=verbose):
@@ -302,8 +314,6 @@ def configure_replica(stanza, pg1_path, pg1_host, pg1_port, pg1_user):
     changes = {
         "hot_standby": "on",
         "primary_conninfo": primary_conninfo,
-        "port": pg1_port,
-        "log_directory": os.path.join(pg1_path, "log"),
         "archive_command": "cd .",
         "archive_mode": "on"
     }
@@ -396,12 +406,12 @@ def modify_hba_conf():
     }]
     util.update_pg_hba_conf(pgV(), new_rules)
 
-def modify_postgresql_conf(stanza, pg1_path, repo1_path, repo1_type):
+def modify_postgresql_conf(stanza, pg1_path, repo1_path, repo1_type, repo1_cipher_type="aes-256-cbc"):
     """Modify 'postgresql.conf' to integrate with pgbackrest."""
     aCmd = (
         f"pgbackrest --stanza={stanza} --pg1-path={pg1_path} "
         f"--repo1-type={repo1_type} --repo1-path={repo1_path} "
-        f"--repo1-cipher-type=aes-256-cbc archive-push %p"
+        f"--repo1-cipher-type={repo1_cipher_type} archive-push %p"
     )
     util.change_pgconf_keyval(pgV(), "archive_command", aCmd, p_replace=True)
     util.change_pgconf_keyval(pgV(), "archive_mode", "on", p_replace=True)
@@ -518,12 +528,13 @@ def update_config(verbose=True):
         "pitr": pitr,
         "create-stanza": create_stanza,
         "create-replica": create_replica,
-        "configure_replica": configure_replica,
+        "configure-replica": configure_replica,
         "list-backups": list_backups,
         "show-config": show_config,
         "write-config": write_config,
         "update-config": update_config, 
-        "set_hbaconf": modify_hba_conf,
-        "set_postgresqlconf": modify_postgresql_conf,
+        "set-hbaconf": modify_hba_conf,
+        "set-postgresqlconf": modify_postgresql_conf,
+        "cleanup-replica": cleanup_replica,
         "command": run_external_command,
     })

src/backrest/install-backrest.py

@@ -2,12 +2,7 @@
 #     Copyright (c)  2022-2025 PGEDGE  #
 
 import os
-import subprocess
-import time
 import sys
-import getpass
-from crontab import CronTab
-import subprocess
 import util
 
 thisDir = os.path.dirname(os.path.realpath(__file__))
@@ -36,6 +31,7 @@ def osSys(p_input, p_display=False):
         util.message("# " + p_input)
     rc = os.system(p_input)
     return rc
+
 def configure_backup_settings():
     """Configure and return the pgBackRest settings."""
     stanza = "pg16"
@@ -93,94 +89,12 @@ def setup_pgbackrest_links():
     osSys("sudo mkdir -p -m 770 /var/lib/pgbackrest")
     osSys(f"sudo chown {usrUsr} -R /var/lib/pgbackrest")
 
-def modify_hba_conf():
-  new_rules = [
-      {
-            "type": "host",
-            "database": "replication",
-            "user": "all",
-            "address": "127.0.0.1/0",
-            "method": "trust"
-      }
-  ]
-  util.update_pg_hba_conf(pgV(), new_rules)
-
-def create_or_update_job(crontab_lines, job_comment, detailed_comment, new_job):
-    job_identifier = f"# {job_comment}"
-    detailed_comment_line = f"# {detailed_comment}\n"
-    job_exists = False
-
-    for i, line in enumerate(crontab_lines):
-        if job_identifier in line:
-            crontab_lines[i] = job_identifier + "\n"
-            if i + 1 < len(crontab_lines):
-                crontab_lines[i + 1] = detailed_comment_line
-                crontab_lines[i + 2] = new_job
-            job_exists = True
-            break
-
-    if not job_exists:
-        crontab_lines.extend([job_identifier + "\n", detailed_comment_line, new_job])
-
-def define_cron_job():
-    stanza = util.get_value("BACKUP", "stanza")
-    full_backup_command = f"pgbackrest --stanza={stanza} --type=full backup"
-    incr_backup_command = f"pgbackrest --stanza={stanza} --type=incr backup"
-    expire_backup_command = f"pgbackrest --stanza={stanza} expire"
-
-    run_as_user = 'root'
-
-    # Crontab entries with detailed comments
-    full_backup_cron = f"0 1 * * * {run_as_user} {full_backup_command}\n"
-    incr_backup_cron = f"0 * * * * {run_as_user} {incr_backup_command}\n"
-    expire_backup_cron = f"30 1 * * * {run_as_user} {expire_backup_command}\n"
-
-    # Detailed comments for each job
-    full_backup_comment = "Performs a full backup daily at 1 AM."
-    incr_backup_comment = "Performs an incremental backup every hour."
-    expire_backup_comment = "Manages backup retention, expiring old backups at 1:30 AM daily."
-
-    system_crontab_path = "/etc/crontab"
-    backrest_crontab_path = "backrest.crontab"
-
-    with open(system_crontab_path, 'r') as file:
-        existing_crontab = file.readlines()
-
-    create_or_update_job(existing_crontab, "FullBackup", full_backup_comment, full_backup_cron)
-    create_or_update_job(existing_crontab, "IncrementalBackup", incr_backup_comment, incr_backup_cron)
-    create_or_update_job(existing_crontab, "ExpireBackup", expire_backup_comment, expire_backup_cron)
-
-    with open(backrest_crontab_path, 'w') as file:
-        file.writelines(existing_crontab)
-
-    osSys(f"sudo cat {backrest_crontab_path} | sudo tee {system_crontab_path} > /dev/null", False)
-
-def fetch_backup_config():
-    """Fetch and return the pgBackRest configuration from system settings."""
-    config = {}
-    params = [
-        "stanza", "restore_path", "backup-type", "repo1-retention-full",
-        "repo1-retention-full-type", "repo1-path", "repo1-host-user",
-        "repo1-host", "repo1-cipher-type", "repo1-cipher-pass", "repo1-s3-bucket",
-        "repo1-s3-key-secret", "repo1-s3-key", "repo1-s3-region", "repo1-s3-endpoint",
-        "log-level-console", "repo1-type", "process-max", "compress-level",
-        "pg1-path", "pg1-user", "pg1-database", "db-socket-path", "pg1-port",
-        "pg1-host"
-    ]
-
-    # Fetch all parameters
-    for param in params:
-        config[param] = util.get_value("BACKUP", param)
-
-    return config
-
 def print_header(header):
     bold_start = "\033[1m"
     bold_end = "\033[0m"
     print(bold_start + "##### " + header + " #####"+ bold_end)
 
 def main():
-    stanza = pgV()
     if os.path.isdir(f"/var/lib/pgbackrest/"):
         util.message("/var/lib/pgbackrest directory already exists")
 

src/conf/versions.sql

@@ -1,7 +1,7 @@
 
 DROP TABLE IF EXISTS hub;
 CREATE TABLE hub(v TEXT NOT NULL PRIMARY KEY, c TEXT NOT NULL, d TEXT NOT NULL);
-INSERT INTO hub VALUES ('25.0.0-alpha3', 'Constellation',  '20250421');
+INSERT INTO hub VALUES ('25.0.0-alpha4', 'Constellation',  '20250421');
 
 DROP VIEW  IF EXISTS v_versions;
 DROP VIEW  IF EXISTS v_products;