GitLab 在低配置服务器中运行 502 异常

Question

我在/etc/gitlab/gitlab.rb中进行了如下配置，但奇怪的是GitLab以非常有规律的方式出现502，大约每10分钟出现一次

external_url 'https://example.com'
gitlab_rails['trusted_proxies'] = ['127.0.0.1','localhost']
gitlab_rails['terraform_state_enabled'] = false
gitlab_rails['usage_ping_enabled'] = false
gitlab_rails['redis_host'] = "127.0.0.1"
gitlab_rails['redis_port'] = 6379
gitlab_rails['sentry_enabled'] = false
gitlab_workhorse['listen_network'] = "tcp"
gitlab_workhorse['listen_addr'] = "127.0.0.1:7991"
puma['worker_processes'] = 0
puma['min_threads'] = 1
puma['max_threads'] = 3
puma['port'] = 7992
puma['per_worker_max_memory_mb'] = 512
sidekiq['max_concurrency'] = 5
postgresql['shared_buffers'] = "512MB"
postgresql['max_worker_processes'] = 2
redis['enable'] = false
nginx['enable'] = false
gitlab_ci['gitlab_ci_all_broken_builds'] = false
gitlab_ci['gitlab_ci_add_pusher'] = false
gitlab_rails['gitlab_kas_enabled'] = false
gitlab_kas['enable'] = false
mattermost['enable'] = false
mattermost_nginx['enable'] = false
prometheus_monitoring['enable'] = false
gitlab_rails['kerberos_enabled'] = false
sentinel['enable'] = false

这是puma日志，看起来它只是不断重启

{"timestamp":"2023-09-15T06:07:21.400Z","pid":2590149,"message":"- Gracefully stopping, waiting for requests to finish"}
=== puma startup: 2023-09-15 14:08:15 +0800 ===
{"timestamp":"2023-09-15T06:13:54.109Z","pid":2591848,"message":"- Gracefully stopping, waiting for requests to finish"}
=== puma startup: 2023-09-15 14:14:32 +0800 ===
{"timestamp":"2023-09-15T06:20:10.861Z","pid":2593485,"message":"- Gracefully stopping, waiting for requests to finish"}
=== puma startup: 2023-09-15 14:20:48 +0800 ===

由于我的配置很低，默认配置无法正常启动，怎么办？我需要大家的帮助，谢谢。

Answer 1

我们的 GitLab 服务器上出现了这个问题，在这个问题中进行了解释。

问题出在我们的

gitlab.rb

文件中的这一行。

puma['worker_processes'] = 0

只需将其删除（恢复为默认值）即可解决我们的崩溃问题！