背景
immich 使用 postgres 作为数据库，发现无法打开相关服务，仔细查看是无法连接数据库，数据库不断重启。docker log内容为
2024-12-16 20:54:26 2024-12-16 12:54:26.288 UTC [1] LOG: redirecting log output to logging collector process
2024-12-16 20:54:26 2024-12-16 12:54:26.288 UTC [1] HINT: Future log output will appear in directory "log".
2024-12-16 20:50:12
2024-12-16 20:50:12 PostgreSQL Database directory appears to contain a database; Skipping initialization
2024-12-16 20:50:12
2024-12-16 20:50:46
没有全部的日志内容，需要去查看 log 文件夹下对应时间的详细日志
[2024-12-16T12:54:39Z INFO service::utils::clean] Find directory "pg_vectors/indexes/17332/segments/d5f94e4c-3dca-4426-91c8-e401381e72b3".
2024-12-16 12:54:42.201 UTC [46] FATAL: the database system is starting up
2024-12-16 12:54:42.213 UTC [47] FATAL: the database system is starting up
2024-12-16 12:54:45.206 UTC [48] FATAL: the database system is starting up
2024-12-16 12:54:45.221 UTC [49] FATAL: the database system is starting up
2024-12-16 12:54:47.859 UTC [28] LOG: database system was not properly shut down; automatic recovery in progress
FATAL: could not create file "base/197420/989335": File exists
结果
分析上面日志，可以看到有文件占用，导致无法创建对应文件 重点见
FATAL: could not create file "base/197420/989335": File exists
- 如果将 989335 文件删除，重启 postgres 数据库。结果 无效果，数据库依然重启。
- 将文件按照文件名进行排列，可以看到 989335 以及 989335_init ，可以将类似 xx_init 一起删除。等待 postgresl 数据库运行。后续 数据库启动成功，服务正常。