Перейти к основному содержимому

Задание 1: Администрирование кластера

В этом разделе будут отработаны навыки по управлению кластером, включая получение информации о его состоянии, а также остановку и запуск узлов с проверкой статуса etcd.

Вывод состояния кластера​

Определите статус сервиса pangolin-manager на узлах srv1 и srv2, используя возможности systemd:

[student@srv1 ~]$ for comp in srv{1,2}; do echo -e "\n$comp:"; ssh $comp sudo systemctl status pangolin-manager --no-pager; done
Вывод команды:

srv1:
● pangolin-manager.service - Runners to orchestrate a high-availability PostgreSQL
     Loaded: loaded (/usr/lib/systemd/system/pangolin-manager.service; enabled; preset: disabled)
     Active: active (running) since Tue 2026-09-08 07:56:42 UTC; 23h ago
    Process: 184249 ExecStartPre=/bin/mkdir -p /var/run/postgresql (code=exited, status=0/SUCCESS)
    Process: 184250 ExecStartPre=/bin/chown -R postgres:postgres /var/run/postgresql (code=exited, status=0/SUCCESS)
   Main PID: 184251 (pangolin-manage)
      Tasks: 18 (limit: 22804)
     Memory: 355.1M (peak: 359.7M)
        CPU: 3min 44.243s
     CGroup: /system.slice/pangolin-manager.service
             ├─184251 /opt/pangolin-manager/bin/pangolin-manager-bin/pangolin-manager.bin /etc/pangolin-manager/postgres.yml
             ├─186351 /usr/pangolin-6.7/bin/postgres -D /pgdata/06/data/ --config-file=/pgdata/06/data/postgresql.conf --listen_addresses=0.0.0.0 --port=5433 --cluster_name=Pobeda --wal_level=replica --hot_standby=on --max_connections=111 --max_wal_senders=10 --max_prepared_transactions=6 --max_locks_per_transaction=64 --track_commit_timestamp=False --max_replication_slots=10 --max_worker_processes=32 --wal_log_hints=True --authentication_max_workers=16 --enable_page_compression=off
             ├─186354 "postgres: Pobeda: logger "
             ├─186355 "postgres: Pobeda: checkpointer "
             ├─186356 "postgres: Pobeda: background writer "
             ├─186357 "postgres: Pobeda: startup recovering 00000007000000000000000E"
             ├─186358 "postgres: Pobeda: idle sessions terminator "
             ├─186359 "postgres: Pobeda: walreceiver streaming 0/F000000"
             ├─186360 "postgres: Pobeda: integrity check launcher "
             ├─186361 "postgres: Pobeda: license checker "
             ├─186362 "postgres: Pobeda: authproc "
             ├─186363 "postgres: Pobeda: password policy cache "
             └─186368 "postgres: Pobeda: patroni postgres 127.0.0.1(52492) idle"

srv2:
● pangolin-manager.service - Runners to orchestrate a high-availability PostgreSQL
     Loaded: loaded (/usr/lib/systemd/system/pangolin-manager.service; enabled; preset: disabled)
     Active: active (running) since Mon 2026-09-07 12:25:04 UTC; 1 day 18h ago
   Main PID: 91425 (pangolin-manage)
      Tasks: 21 (limit: 22804)
     Memory: 488.5M (peak: 492.2M)
        CPU: 6min 53.224s
     CGroup: /system.slice/pangolin-manager.service
             ├─ 91425 /opt/pangolin-manager/bin/pangolin-manager-bin/pangolin-manager.bin /etc/pangolin-manager/postgres.yml
             ├─105092 /usr/pangolin-6.7/bin/postgres -D /pgdata/06/data/ --config-file=/pgdata/06/data/postgresql.conf --listen_addresses=0.0.0.0 --port=5433 --cluster_name=Pobeda --wal_level=replica --hot_standby=on --max_connections=111 --max_wal_senders=10 --max_prepared_transactions=6 --max_locks_per_transaction=64 --track_commit_timestamp=False --max_replication_slots=10 --max_worker_processes=32 --wal_log_hints=True --authentication_max_workers=16 --enable_page_compression=off
             ├─105095 "postgres: Pobeda: logger "
             ├─105096 "postgres: Pobeda: checkpointer "
             ├─105097 "postgres: Pobeda: background writer "
             ├─105099 "postgres: Pobeda: idle sessions terminator "
             ├─105100 "postgres: Pobeda: walwriter "
             ├─105101 "postgres: Pobeda: autovacuum launcher "
             ├─105102 "postgres: Pobeda: integrity check launcher "
             ├─105103 "postgres: Pobeda: license checker "
             ├─105104 "postgres: Pobeda: authproc "
             ├─105105 "postgres: Pobeda: password policy cache "
             ├─105106 "postgres: Pobeda: pg_cron launcher "
             ├─105107 "postgres: Pobeda: logical replication launcher "
             ├─105112 "postgres: Pobeda: patroni postgres 127.0.0.1(52574) idle"
             └─105114 "postgres: Pobeda: walsender patroni <IP-Address1>(40430) streaming 0/F000000"

Убедитесь, что на арбитре srv3 сервис pangolin-manager не запущен:

[student@srv1 ~]$ ssh srv3 sudo systemctl status pangolin-manager --no-pager
Unit pangolin-manager.service could not be found.

Получите список узлов в кластере физической репликации средствами автоматизации управлением кластера:

[student@srv1 ~]$ sudo su - postgres
===============================================================
 You are connected to Pangolin Cluster (Pobeda)
 - Hostname             : srv1
 - IP Address           : <IP-Address1>
---------------------------------------------------------------
 - node ( master )      : <IP-Address2>
 - node ( replica )     : <IP-Address1>
 - arbiter              : <IP-Address3>
 - Manager version      : 2.1.13
 - Pangolin version     : 6.7.0
===============================================================
[postgres@srv1 ~]$ list
+ Cluster: Pobeda (7683495148262105741) ------------+-----------+----+-----------+
| Member        | Host               | Role         | State     | TL | Lag in MB |
+---------------+--------------------+--------------+-----------+----+-----------+
| <IP-Address1> | <IP-Address1>:5433 | Sync Standby | streaming |  5 |         0 |
| <IP-Address2> | <IP-Address2>:5433 | Leader       | running   |  5 |           |
+---------------+--------------------+--------------+-----------+----+-----------+

Остановка и запуск узла кластера​

Остановите реплику srv1:

[student@srv1 ~]$ sudo systemctl stop pangolin-manager --no-pager
[student@srv1 ~]$ sudo systemctl status pangolin-manager --no-pager
○ pangolin-manager.service - Runners to orchestrate a high-availability PostgreSQL
     Loaded: loaded (/usr/lib/systemd/system/pangolin-manager.service; enabled; preset: disabled)
     Active: inactive (dead) since Tue 2026-09-08 12:09:47 UTC; 2s ago
   Duration: 20min 54.719s
    Process: 275706 ExecStartPre=/bin/mkdir -p /var/run/postgresql (code=exited, status=0/SUCCESS)
    Process: 275707 ExecStartPre=/bin/chown -R postgres:postgres /var/run/postgresql (code=exited, status=0/SUCCESS)
    Process: 275708 ExecStart=/opt/pangolin-manager/bin/pangolin-manager-bin/pangolin-manager.bin /etc/pangolin-manager/postgres.yml (code=exited, status=0/SUCCESS)
   Main PID: 275708 (code=exited, status=0/SUCCESS)
        CPU: 6.332s
...

Проверьте, сколько сейчас узлов в кластере:

[student@srv1 ~]$ sudo su - postgres
===============================================================
 You are connected to Pangolin Cluster (Pobeda)
 - Hostname             : srv1
 - IP Address           : <IP-Address1>
---------------------------------------------------------------
 - node ( master )      :
 - node ( replica )     :
 - arbiter              : <IP-Address3>
 - Manager version      : 2.1.13
 - Pangolin version     : 6.7.0
===============================================================
[postgres@srv1 ~]$ list
+ Cluster: Pobeda (7683495148262105741) -------+---------+----+-----------+
| Member        | Host               | Role    | State   | TL | Lag in MB |
+---------------+--------------------+---------+---------+----+-----------+
| <IP-Address1> | <IP-Address1>:5433 | Replica | stopped |    |   unknown |
| <IP-Address2> | <IP-Address2>:5433 | Leader  | running |  5 |           |
+---------------+--------------------+---------+---------+----+-----------+
[postgres@srv1 ~]$ exit

Снова запустите реплику srv1:

[student@srv1 ~]$ sudo systemctl start pangolin-manager --no-pager
[student@srv1 ~]$ sudo systemctl status pangolin-manager --no-pager
● pangolin-manager.service - Runners to orchestrate a high-availability PostgreSQL
     Loaded: loaded (/usr/lib/systemd/system/pangolin-manager.service; enabled; preset: disabled)
     Active: active (running) since Wed 2026-09-09 12:12:25 UTC; 6s ago
    Process: 276083 ExecStartPre=/bin/mkdir -p /var/run/postgresql (code=exited, status=0/SUCCESS)
    Process: 276084 ExecStartPre=/bin/chown -R postgres:postgres /var/run/postgresql (code=exited, status=0/SUCCESS)
   Main PID: 276085 (pangolin-manage)
      Tasks: 18 (limit: 22804)
     Memory: 138.1M (peak: 140.1M)
        CPU: 1.135s
     CGroup: /system.slice/pangolin-manager.service
             ├─276085 /opt/pangolin-manager/bin/pangolin-manager-bin/pangolin-manager.bin /etc/pangolin-manager/postgres.yml
             ├─276100 /usr/pangolin-6.7/bin/postgres -D /pgdata/06/data/ --config-file=/pgdata/06/data/postgresql.conf --listen_addresses=0.0.0.0 --port=5433 --cluster_name=Pobeda --wal_level=repli…
             ├─276103 "postgres: Pobeda: logger "
             ├─276104 "postgres: Pobeda: checkpointer "
             ├─276105 "postgres: Pobeda: background writer "
             ├─276106 "postgres: Pobeda: startup recovering 000000050000000000000009"
             ├─276107 "postgres: Pobeda: idle sessions terminator "
             ├─276108 "postgres: Pobeda: authproc "
             ├─276109 "postgres: Pobeda: password policy cache "
             ├─276110 "postgres: Pobeda: walreceiver streaming 0/A000000"
             ├─276113 "postgres: Pobeda: integrity check launcher "
             ├─276114 "postgres: Pobeda: license checker "
             └─276117 "postgres: Pobeda: patroni postgres 127.0.0.1(39610) idle"

После успешного старта реплики проверьте узлы кластера:

[student@srv1 ~]$ sudo su - postgres
===============================================================
 You are connected to Pangolin Cluster (Pobeda)
 - Hostname             : srv1
 - IP Address           : <IP-Address1>
---------------------------------------------------------------
 - node ( master )      : <IP-Address2>
 - node ( replica )     : <IP-Address1>
 - arbiter              : <IP-Address3>
 - Manager version      : 2.1.13
 - Pangolin version     : 6.7.0
===============================================================
[postgres@srv1 ~]$ list
+ Cluster: Pobeda (7683495148262105741) ------------+-----------+----+-----------+
| Member        | Host               | Role         | State     | TL | Lag in MB |
+---------------+--------------------+--------------+-----------+----+-----------+
| <IP-Address1> | <IP-Address1>:5433 | Sync Standby | streaming |  5 |         0 |
| <IP-Address2> | <IP-Address2>:5433 | Leader       | running   |  5 |           |
+---------------+--------------------+--------------+-----------+----+-----------+
[postgres@srv1 ~]$ exit

Подведем итоги​

Вопрос 1

Вопрос 1: Приведен фрагмент списка процессов СУБД Pangolin на одном из узлов кластера, собранного из троих узлов:

             ├─37403 "postgres: Pobeda: logger "
             ├─37404 "postgres: Pobeda: checkpointer "
             ├─37405 "postgres: Pobeda: background writer "
             ├─37406 "postgres: Pobeda: startup recovering 000000040000000000000008"
             ├─37407 "postgres: Pobeda: idle sessions terminator "
             ├─37408 "postgres: Pobeda: authproc "
             ├─37409 "postgres: Pobeda: password policy cache "
             ├─37413 "postgres: Pobeda: integrity check launcher "
             ├─37414 "postgres: Pobeda: license checker "

На каких узлах этот список НЕ мог быть получен?

Вопрос 2

Вопрос 2: Как получить состояние сервиса pangolin-manager.service с помощью systemctl без запуска пэйджера?