# 尝试恢复安装

**URL:** <https://meta.discourse.org/t/trying-to-recover-an-installation/262008>\
**Category:** Self-hosting\
**Created:** [2023年四月17日 21:12 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008 "2023-04-17T21:12:45Z")\
**Posts on this page:** 9\
**Page:** 1

<div class="post-metadata">

**Author:** ![Norike](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/norike/32/299043_2.png) [@Norike](https://meta.discourse.org/u/Norike)\
**Post date:** [2023年四月17日 21:12 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/1 "2023-04-17T21:12:45Z")

</div>

我在 CentOS 机器上按照指南 [此处](https://github.com/discourse/discourse/blob/main/docs/INSTALL-cloud.md) 安装了 Discourse。  
如果这有帮助，我们还使用了 [NGINX Proxy Manager](https://meta.discourse.org/t/use-nginx-proxy-manager-to-manage-multiple-sites-with-discourse/206344)。

安装工作了几周，直到我们不得不重启机器。然后它就无法启动了。这是 `rebuild` 的输出：

```plaintext
Ensuring launcher is up to date
Launcher is up-to-date
2.0.20230409-0052: Pulling from discourse/base
Digest: sha256:dd75ceb9322f79629f8b0bf78cfb0f79ad6bb366b7ead3e1cd32dcb8712ec46f
Status: Image is up to date for discourse/base:2.0.20230409-0052
docker.io/discourse/base:2.0.20230409-0052
/usr/local/lib/ruby/gems/3.2.0/gems/pups-1.1.1/lib/pups.rb
/usr/local/bin/pups --stdin
I, [2023-04-13T09:55:28.002076 #1] INFO -- : Reading from stdin
I, [2023-04-13T09:55:28.005509 #1] INFO -- : > locale-gen $LANG && update-locale
I, [2023-04-13T09:55:28.031953 #1] INFO -- : Generating locales (this might take a while)...
Generation complete.

I, [2023-04-13T09:55:28.032080 #1] INFO -- : > mkdir -p /shared/postgres_run
I, [2023-04-13T09:55:28.034322 #1] INFO -- :
I, [2023-04-13T09:55:28.034470 #1] INFO -- : > chown postgres:postgres /shared/postgres_run
I, [2023-04-13T09:55:28.036190 #1] INFO -- :
I, [2023-04-13T09:55:28.036297 #1] INFO -- : > chmod 775 /shared/postgres_run
I, [2023-04-13T09:55:28.037819 #1] INFO -- :
I, [2023-04-13T09:55:28.037918 #1] INFO -- : > rm -fr /var/run/postgresql
I, [2023-04-13T09:55:28.039663 #1] INFO -- :
I, [2023-04-13T09:55:28.039770 #1] INFO -- : > ln -s /shared/postgres_run /var/run/postgresql
I, [2023-04-13T09:55:28.041327 #1] INFO -- :
I, [2023-04-13T09:55:28.041457 #1] INFO -- : > socat /dev/null UNIX-CONNECT:/shared/postgres_run/.s.PGS
QL.5432 || exit 0 && echo postgres already running stop container ; exit 1
2023/04/13 09:55:28 socat[19] E connect(6, AF=1 "/shared/postgres_run/.s.PGSQL.5432", 36): No such file
or directory
I, [2023-04-13T09:55:28.045222 #1] INFO -- :
I, [2023-04-13T09:55:28.045329 #1] INFO -- : > rm -fr /shared/postgres_run/.s*
I, [2023-04-13T09:55:28.047498 #1] INFO -- :
I, [2023-04-13T09:55:28.047590 #1] INFO -- : > rm -fr /shared/postgres_run/*.pid
I, [2023-04-13T09:55:28.049691 #1] INFO -- :
I, [2023-04-13T09:55:28.049815 #1] INFO -- : > mkdir -p /shared/postgres_run/13-main.pg_stat_tmp
I, [2023-04-13T09:55:28.051658 #1] INFO -- :
I, [2023-04-13T09:55:28.051800 #1] INFO -- : > chown postgres:postgres /shared/postgres_run/13-main.pg_
stat_tmp
I, [2023-04-13T09:55:28.053477 #1] INFO -- :
I, [2023-04-13T09:55:28.057301 #1] INFO -- : File > /etc/service/postgres/run chmod: +x chown:
I, [2023-04-13T09:55:28.060969 #1] INFO -- : File > /etc/service/postgres/log/run chmod: +x chown:
I, [2023-04-13T09:55:28.064663 #1] INFO -- : File > /etc/runit/3.d/99-postgres chmod: +x chown:
I, [2023-04-13T09:55:28.068319 #1] INFO -- : File > /root/upgrade_postgres chmod: +x chown:
I, [2023-04-13T09:55:28.068510 #1] INFO -- : > chown -R root /var/lib/postgresql/13/main
I, [2023-04-13T09:55:37.064277 #1] INFO -- :
I, [2023-04-13T09:55:37.064575 #1] INFO -- : > [! -e /shared/postgres_data] && install -d -m 0755 -o
I, [2023-04-13T09:55:37.066210 #1] INFO -- :
I, [2023-04-13T09:55:37.066255 #1] INFO -- : > chown -R postgres:postgres /shared/postgres_data
I, [2023-04-13T09:55:37.076706 #1] INFO -- :
I, [2023-04-13T09:55:37.076786 #1] INFO -- : > chown -R postgres:postgres /var/run/postgresql
I, [2023-04-13T09:55:37.078711 #1] INFO -- :
I, [2023-04-13T09:55:37.078822 #1] INFO -- : > /root/upgrade_postgres
I, [2023-04-13T09:55:37.082200 #1] INFO -- :
I, [2023-04-13T09:55:37.082297 #1] INFO -- : > rm /root/upgrade_postgres
I, [2023-04-13T09:55:37.083919 #1] INFO -- :
I, [2023-04-13T09:55:37.084109 #1] INFO -- : Replacing data_directory = '/var/lib/postgresql/13/main' w
I, [2023-04-13T09:55:37.084484 #1] INFO -- : Replacing (?-mix:#?listen_addresses *=.*) with listen_addr
I, [2023-04-13T09:55:37.085049 #1] INFO -- : Replacing (?-mix:#?synchronous_commit *=.*) with synchrono
I, [2023-04-13T09:55:37.085423 #1] INFO -- : Replacing (?-mix:#?shared_buffers *=.*) with shared_buffer
I, [2023-04-13T09:55:37.085857 #1] INFO -- : Replacing (?-mix:#?work_mem *=.*) with work_mem = $db_work
I, [2023-04-13T09:55:37.086303 #1] INFO -- : Replacing (?-mix:#?default_text_search_config *=.*) with d
I, [2023-04-13T09:55:37.086614 #1] INFO -- : > install -d -m 0755 -o postgres -g postgres /shared/postg
I, [2023-04-13T09:55:37.089011 #1] INFO -- :
I, [2023-04-13T09:55:37.089245 #1] INFO -- : Replacing (?-mix:#?checkpoint_segments *=.*) with checkpoi
I, [2023-04-13T09:55:37.089516 #1] INFO -- : Replacing (?-mix:#?logging_collector *=.*) with logging_co
I, [2023-04-13T09:55:37.089887 #1] INFO -- : Replacing (?-mix:#?log_min_duration_statement *=.*) with l
I, [2023-04-13T09:55:37.090322 #1] INFO -- : Replacing (?-mix:^#local +replication +postgres +peer$) wi
I, [2023-04-13T09:55:37.090509 #1] INFO -- : Replacing (?-mix:^host.*all.*all.*127.*$) with host all al
I, [2023-04-13T09:55:37.090946 #1] INFO -- : Replacing (?-mix:^host.*all.*all.*::1\/128.*$) with host a
I, [2023-04-13T09:55:37.091279 #1] INFO -- : > HOME=/var/lib/postgresql USER=postgres exec chpst -u pos
I, [2023-04-13T09:55:37.092562 #1] INFO -- : > sleep 5
2023-04-13 09:55:37.239 UTC [42] LOG: starting PostgreSQL 13.10 (Debian 13.10-1.pgdg110+1) on x86_64-pc
2023-04-13 09:55:37.240 UTC [42] LOG: listening on IPv4 address "0.0.0.0", port 5432
2023-04-13 09:55:37.240 UTC [42] LOG: listening on IPv6 address "::", port 5432
2023-04-13 09:55:37.287 UTC [42] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
2023-04-13 09:55:37.338 UTC [45] LOG: database system was interrupted; last known up at 2023-04-13 08:3
2023-04-13 09:55:37.795 UTC [45] LOG: invalid resource manager ID in primary checkpoint record
2023-04-13 09:55:37.795 UTC [45] PANIC: could not locate a valid checkpoint record
2023-04-13 09:55:38.256 UTC [42] LOG: startup process (PID 45) was terminated by signal 6: Aborted
2023-04-13 09:55:38.256 UTC [42] LOG: aborting startup due to startup process failure
2023-04-13 09:55:38.283 UTC [42] LOG: database system is shut down
I, [2023-04-13T09:55:42.094474 #1] INFO -- :
I, [2023-04-13T09:55:42.094602 #1] INFO -- : > su postgres -c 'createdb discourse' || true
createdb: error: could not connect to database template1: connection to server on socket "/var/run/postg
        Is the server running locally and accepting connections on that socket?
I, [2023-04-13T09:55:42.130700 #1] INFO -- :
I, [2023-04-13T09:55:42.130828 #1] INFO -- : > su postgres -c 'psql discourse -c "create user discourse
psql: error: connection to server on socket "/var/run/postgresql/.s.PGSQL.5432" failed: No such file or directory
        Is the server running locally and accepting connections on that socket?
I, [2023-04-13T09:55:42.166116 #1] INFO -- :
I, [2023-04-13T09:55:42.166242 #1] INFO -- : > su postgres -c 'psql discourse -c "grant all privileges on database discourse to discourse;"' || true
psql: error: connection to server on socket "/var/run/postgresql/.s.PGSQL.5432" failed: No such file or directory
        Is the server running locally and accepting connections on that socket?
I, [2023-04-13T09:55:42.201461 #1] INFO -- :
I, [2023-04-13T09:55:42.201617 #1] INFO -- : > su postgres -c 'psql discourse -c "alter schema public owner to discourse;"'
psql: error: connection to server on socket "/var/run/postgresql/.s.PGSQL.5432" failed: No such file or directory
        Is the server running locally and accepting connections on that socket?
I, [2023-04-13T09:55:42.236768 #1] INFO -- :
I, [2023-04-13T09:55:42.236996 #1] INFO -- : Terminating async processes

FAILED
--------------------
Pups::ExecError: su postgres -c 'psql discourse -c "alter schema public owner to discourse;"' failed with return #<Process::Status: pid 55 exit 2>
Location of failure: /usr/local/lib/ruby/gems/3.2.0/gems/pups-1.1.1/lib/pups/exec_command.rb:117:in `spawn'
exec failed with the params "su postgres -c 'psql $db_name -c \\\"alter schema public owner to $db_user;\\\"'"
bootstrap failed with exit code 2
**FAILED TO BOOTSTRAP** please scroll up and look for earlier error messages, there may be more than one.
./discourse-doctor may help diagnose the problem.
cd04fbb38f1ef61e418d680b969c2c056439e3b1dbfe64608524a8d3361cd91c

```

`discourse-doctor` 无法找到正在运行的 docker 应用程序：

```plaintext
==================== SERIOUS PROBLEM!!!! ====================
app not running!
Attempting to rebuild

```

并尝试重建，因此基本上会以与上面相同的日志结束。

提前感谢！

---

<div class="post-metadata">

**Author:** ![shyguy](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/shyguy/32/295079_2.png) [@shyguy](https://meta.discourse.org/u/shyguy)\
**Post date:** [2023年四月18日 02:06 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/2 "2023-04-18T02:06:57Z")

</div>

> `PANIC: could not locate a valid checkpoint record`

我认为您的数据库已损坏。依我看，最好从备份中恢复。我认为 Discourse 默认会进行备份吗？请查看 `discourse/shared/app/backups` 或其他位置。

* * *

如果您没有 Discourse 备份，那么这更多是一个 PostgreSQL 支持主题，而不是 Discourse 主题。我猜我可以帮忙，但我不是专家。

您可能需要使用这个：

> **[pg\_resetwal](https://www.postgresql.org/docs/13/app-pgresetwal.html)**
>
> pg\_resetwal pg\_resetwal — reset the write-ahead log and other control information of a PostgreSQL database cluster Synopsis pg\_resetwal \[ -f …

---

<div class="post-metadata">

**Author:** ![Ed\_S](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/ed_s/32/134015_2.png) [@Ed\_S](https://meta.discourse.org/u/Ed_S)\
**Post date:** [2023年四月18日 09:09 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/3 "2023-04-18T09:09:43Z")

</div>

> [@shyguy](#):
>
> 我认为你的数据库已损坏。依我看，最好从备份恢复。

听起来确实如此。我希望你能找到解决方案，@Norike……与此同时，我觉得这相当令人担忧——我相信软件重启应该能使数据库保持干净状态。这是硬重启，也就是断电重启吗？

我想知道数据库存储在什么文件系统上是否会有所不同。`df -T` 告诉你什么？

---

<div class="post-metadata">

**Author:** ![Norike](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/norike/32/299043_2.png) [@Norike](https://meta.discourse.org/u/Norike)\
**Post date:** [2023年四月18日 09:24 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/4 "2023-04-18T09:24:19Z")

</div>

谢谢！

确实，我在 `/var/discourse/shared/standalone/backups/default` 文件夹中找到了一些备份，这听起来很有希望。  
但是，根据 [从命令行恢复备份](https://meta.discourse.org/t/restore-a-backup-from-the-command-line/108034) 指南，容器需要运行才能导入数据库，但 `./launcher enter app` 命令失败了：

```plaintext
检测到 x86_64 架构。
警告：containers/app.yml 文件是可被所有人读取的。您可以通过运行以下命令来保护此文件：chmod o-rwx containers/app.yml
守护进程的错误响应：未找到容器：app

```

我猜我应该从干净的安装开始。是否有重新开始的指南？以下是否正确？

- 将备份从 `/var/discourse/shared/standalone/backups/default` 文件夹复制到 `/var/discourse` 文件夹之外的其他位置
- 删除 `/var/discourse` 文件夹
- 按照 [安装云](https://github.com/discourse/discourse/blob/main/docs/INSTALL-cloud.md#5-install-discourse) 指南从头开始安装 Discourse，以获得新的运行实例。
- 使用 GUI 或上述命令行指南恢复备份

---

<div class="post-metadata">

**Author:** ![Norike](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/norike/32/299043_2.png) [@Norike](https://meta.discourse.org/u/Norike)\
**Post date:** [2023年四月18日 09:39 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/5 "2023-04-18T09:39:47Z")

</div>

> [@Ed\_S](#):
>
> 听起来是这样。我希望你能找到一个解决方案，@Norike……与此同时，我觉得这相当令人担忧——我相信软件重启应该能使数据库保持干净的状态。这是硬重启，也就是电源循环吗？

这是在应用了一些更新后进行的软件重启。

> [@Ed\_S](#):
>
> `df -T` 告诉了你什么？

这个：

```plaintext
文件系统 类型 1K-块 已用 可用 已用% 已挂载在
devtmpfs devtmpfs 4096 0 4096 0% /dev
tmpfs tmpfs 3869940 0 3869940 0% /dev/shm
tmpfs tmpfs 1547980 17600 1530380 2% /run
/dev/mapper/cs-root xfs 73364480 16750976 56613504 23% /
/dev/sda2 xfs 1038336 395812 642524 39% /boot
/dev/mapper/cs-home xfs 893122476 10838052 882284424 2% /home
/dev/sda1 vfat 613160 7644 605516 2% /boot/efi
overlay overlay 73364480 16750976 56613504 23% /var/lib/docker/overlay2/a9c2622d4167f08ff4697a4c49febf55a7e460e087c41c05e6c1bbd321b13f62/merged
overlay overlay 73364480 16750976 56613504 23% /var/lib/docker/overlay2/765273939fa75096c994588ba9c9bac6b5a9b909a60a962169b83f7c8a213b7f/merged
tmpfs tmpfs 773988 4 773984 1% /run/user/1000

```

---

<div class="post-metadata">

**Author:** ![Ed\_S](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/ed_s/32/134015_2.png) [@Ed\_S](https://meta.discourse.org/u/Ed_S)\
**Post date:** [2023年四月18日 09:53 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/6 "2023-04-18T09:53:20Z")

</div>

谢谢 - 我不熟悉 XFS，但快速搜索表明它可以使用。也许 `xfs_info /` 会提供有用的信息。

---

<div class="post-metadata">

**Author:** ![shyguy](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/shyguy/32/295079_2.png) [@shyguy](https://meta.discourse.org/u/shyguy)\
**Post date:** [2023年四月18日 15:54 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/7 "2023-04-18T15:54:47Z")

</div>

> [@Norike](#):
>
> 我在 `/var/discourse/shared/standalone/backups/default` 文件夹中找到了一些备份，这听起来很有希望。

好的，然后将整个目录复制到某个地方进行妥善保管：

`/var/discourse/shared`

这样，您就可以尝试从备份恢复，甚至在需要时尝试修复损坏的数据库。

然后删除您的 Docker 容器，运行 `docker image prune -a`，删除 `/var/discourse`，然后重新安装 Discourse。之后将最新的备份文件复制到位并尝试恢复。

---

<div class="post-metadata">

**Author:** ![Norike](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/norike/32/299043_2.png) [@Norike](https://meta.discourse.org/u/Norike)\
**Post date:** [2023年四月18日 21:23 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/8 "2023-04-18T21:23:13Z")

</div>

这恢复了它。谢谢！

---

<div class="post-metadata">

**Author:** ![system](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/system/32/443519_2.png) [@system](https://meta.discourse.org/u/system)\
**Post date:** [2023年五月18日 21:23 UTC](https://meta.discourse.org/t/trying-to-recover-an-installation/262008/9 "2023-05-18T21:23:39Z")

</div>

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.