Multi-site upgrade leaves _some_ sites unavailable

AI-generated summary

A recent multi-site upgrade to Discourse 2.8.0.beta2 has caused some secondary sites to become unavailable, while the primary site and other secondary sites remain operational. hellekin initially suspected the issue might be related to the discourse-calendar plugin, but later discovered it was actually caused by the discourse-prometheus-alert-receiver plugin.

The error log revealed a NoMethodError in lib/topic_query.rb, which seemed to be triggered by the plugin even though it was not enabled for the site. pfaffman suggested trying safe mode and checking themes and plugins, but hellekin found that the issue persisted even in safe mode.

Further investigation revealed that the site would work normally when the discourse-prometheus-alert-receiver plugin was enabled, but would break when it was not enabled. hellekin believes this to be a bug.

With the recent upgrade frenzy, leading to 2.8.0.beta2, something strange happened that I’m currently investigating. The effect is that on a multi-site install, some but not all secondary sites do not respond anymore. The primary site is up, and some secondary sites operate normally.

I suspect it may have to do with the discourse-calendar issue already reported on stable, but that’s just a hunch.

I found this in the log for one of the sites:

error log
[thx_zoethical_org] Started GET "/t/nos-reseaux-tisses/77" for [REDACTED IP] at 2021-06-11 11:19:51 +0000
[thx_zoethical_org] Processing by TopicsController#show as HTML
[thx_zoethical_org]   Parameters: {"slug"=>"nos-reseaux-tisses", "topic_id"=>"77"}
[thx_zoethical_org] Completed 500 Internal Server Error in 388ms (ActiveRecord: 0.0ms | Allocations: 74642)
[thx_zoethical_org] NoMethodError (undefined method `include?' for nil:NilClass)
lib/topic_query.rb:111:in `block in apply_custom_filters'
lib/topic_query.rb:110:in `each'
lib/topic_query.rb:110:in `apply_custom_filters'
lib/topic_query.rb:890:in `default_results'
lib/topic_query.rb:523:in `unread_results'
lib/topic_query.rb:218:in `list_suggested_for'
lib/topic_view.rb:572:in `suggested_topics'
app/serializers/suggested_topics_mixin.rb:16:in `include_suggested_topics?'
(eval):6:in `_fast_attributes'
app/controllers/topics_controller.rb:1158:in `block (2 levels) in perform_show_response'
app/controllers/topics_controller.rb:1153:in `perform_show_response'
app/controllers/topics_controller.rb:162:in `show'
app/controllers/application_controller.rb:395:in `block in with_resolved_locale'
app/controllers/application_controller.rb:395:in `with_resolved_locale'
lib/middleware/omniauth_bypass_middleware.rb:71:in `call'
lib/content_security_policy/middleware.rb:12:in `call'
lib/middleware/anonymous_cache.rb:355:in `call'
config/initializers/008-rack-cors.rb:25:in `call'
config/initializers/100-quiet_logger.rb:23:in `call'
config/initializers/100-silence_logger.rb:31:in `call'
lib/middleware/enforce_hostname.rb:23:in `call'
lib/middleware/request_tracker.rb:177:in `call'

Did you do a git pull first (unnecessary if you do a rebuild rather then a bootstrap).

Another wild guess is that it’s theme related.

Check themes and plugins. Try safe mode?

Thank you @pfaffman. I used rebuild (twice) and safe-mode did not help. I’m currently rebuilding on another machine to avoid multisite and see if restoring from backups work.

Just to add: published pages (under /pub) work normally. Oh, and /admin too… Maybe I can solve it…


Apparently it comes from discourse-prometheus-alert-receiver which is not enabled for this site…

plugins/discourse-prometheus-alert-receiver/plugin.rb:152:in `block (2 levels) in activate!'
lib/topic_query.rb:111:in `block in apply_custom_filters'
lib/topic_query.rb:110:in `each'
lib/topic_query.rb:110:in `apply_custom_filters'
lib/topic_query.rb:890:in `default_results'
lib/topic_query.rb:507:in `latest_results'
lib/topic_query.rb:242:in `list_latest'
app/controllers/categories_controller.rb:51:in `block (2 levels) in index'
actionpack (6.1.3.2) lib/action_controller/metal/mime_responds.rb:214:in `respond_to'
app/controllers/categories_controller.rb:40:in `index'
actionpack (6.1.3.2) lib/action_controller/metal/basic_implicit_render.rb:6:in `send_action'
actionpack (6.1.3.2) lib/abstract_controller/base.rb:228:in `process_action'
actionpack (6.1.3.2) lib/action_controller/metal/rendering.rb:30:in `process_action'
actionpack (6.1.3.2) lib/abstract_controller/callbacks.rb:42:in `block in process_action'
activesupport (6.1.3.2) lib/active_support/callbacks.rb:117:in `block in run_callbacks'
app/controllers/application_controller.rb:395:in `block in with_resolved_locale'
i18n (1.8.10) lib/i18n.rb:314:in `with_locale'
app/controllers/application_controller.rb:395:in `with_resolved_locale'
activesupport (6.1.3.2) lib/active_support/callbacks.rb:126:in `block in run_callbacks'
activesupport (6.1.3.2) lib/active_support/callbacks.rb:137:in `run_callbacks'
actionpack (6.1.3.2) lib/abstract_controller/callbacks.rb:41:in `process_action'
actionpack (6.1.3.2) lib/action_controller/metal/rescue.rb:22:in `process_action'
actionpack (6.1.3.2) lib/action_controller/metal/instrumentation.rb:34:in `block in process_action'
activesupport (6.1.3.2) lib/active_support/notifications.rb:203:in `block in instrument'
activesupport (6.1.3.2) lib/active_support/notifications/instrumenter.rb:24:in `instrument'
activesupport (6.1.3.2) lib/active_support/notifications.rb:203:in `instrument'
actionpack (6.1.3.2) lib/action_controller/metal/instrumentation.rb:33:in `process_action'
actionpack (6.1.3.2) lib/action_controller/metal/params_wrapper.rb:249:in `process_action'
activerecord (6.1.3.2) lib/active_record/railties/controller_runtime.rb:27:in `process_action'
actionpack (6.1.3.2) lib/abstract_controller/base.rb:165:in `process'
actionview (6.1.3.2) lib/action_view/rendering.rb:39:in `process'
rack-mini-profiler (2.3.2) lib/mini_profiler/profiling_methods.rb:111:in `block in profile_method'


actionpack (6.1.3.2) lib/action_controller/metal.rb:190:in `dispatch'
actionpack (6.1.3.2) lib/action_controller/metal.rb:254:in `dispatch'
actionpack (6.1.3.2) lib/action_dispatch/routing/route_set.rb:50:in `dispatch'
actionpack (6.1.3.2) lib/action_dispatch/routing/route_set.rb:33:in `serve'
actionpack (6.1.3.2) lib/action_dispatch/journey/router.rb:50:in `block in serve'
actionpack (6.1.3.2) lib/action_dispatch/journey/router.rb:32:in `each'
actionpack (6.1.3.2) lib/action_dispatch/journey/router.rb:32:in `serve'
actionpack (6.1.3.2) lib/action_dispatch/routing/route_set.rb:842:in `call'
lib/middleware/omniauth_bypass_middleware.rb:71:in `call'
rack (2.2.3) lib/rack/tempfile_reaper.rb:15:in `call'
rack (2.2.3) lib/rack/conditional_get.rb:27:in `call'
rack (2.2.3) lib/rack/head.rb:12:in `call'
actionpack (6.1.3.2) lib/action_dispatch/http/permissions_policy.rb:22:in `call'
lib/content_security_policy/middleware.rb:12:in `call'
lib/middleware/anonymous_cache.rb:355:in `call'
config/initializers/008-rack-cors.rb:25:in `call'
rack (2.2.3) lib/rack/session/abstract/id.rb:266:in `context'
rack (2.2.3) lib/rack/session/abstract/id.rb:260:in `call'
actionpack (6.1.3.2) lib/action_dispatch/middleware/cookies.rb:689:in `call'
actionpack (6.1.3.2) lib/action_dispatch/middleware/callbacks.rb:27:in `block in call'
activesupport (6.1.3.2) lib/active_support/callbacks.rb:98:in `run_callbacks'
actionpack (6.1.3.2) lib/action_dispatch/middleware/callbacks.rb:26:in `call'
actionpack (6.1.3.2) lib/action_dispatch/middleware/actionable_exceptions.rb:18:in `call'
actionpack (6.1.3.2) lib/action_dispatch/middleware/debug_exceptions.rb:29:in `call'
actionpack (6.1.3.2) lib/action_dispatch/middleware/show_exceptions.rb:33:in `call'
logster (2.9.6) lib/logster/middleware/reporter.rb:43:in `call'
railties (6.1.3.2) lib/rails/rack/logger.rb:37:in `call_app'
railties (6.1.3.2) lib/rails/rack/logger.rb:28:in `call'
config/initializers/100-quiet_logger.rb:23:in `call'
config/initializers/100-silence_logger.rb:31:in `call'
actionpack (6.1.3.2) lib/action_dispatch/middleware/remote_ip.rb:81:in `call'
actionpack (6.1.3.2) lib/action_dispatch/middleware/request_id.rb:26:in `call'
lib/middleware/enforce_hostname.rb:23:in `call'
rack (2.2.3) lib/rack/method_override.rb:24:in `call'
rack (2.2.3) lib/rack/sendfile.rb:110:in `call'
actionpack (6.1.3.2) lib/action_dispatch/middleware/host_authorization.rb:92:in `call'
plugins/discourse-prometheus/lib/middleware/metrics.rb:17:in `call'
rack-mini-profiler (2.3.2) lib/mini_profiler/profiler.rb:384:in `call'


message_bus (3.3.6) lib/message_bus/rack/middleware.rb:61:in `call'
lib/middleware/request_tracker.rb:177:in `call'
rails_multisite (3.0.0) lib/rails_multisite/middleware.rb:25:in `call'
railties (6.1.3.2) lib/rails/engine.rb:539:in `call'
railties (6.1.3.2) lib/rails/railtie.rb:207:in `public_send'
railties (6.1.3.2) lib/rails/railtie.rb:207:in `method_missing'
rack (2.2.3) lib/rack/urlmap.rb:74:in `block in call'
rack (2.2.3) lib/rack/urlmap.rb:58:in `each'
rack (2.2.3) lib/rack/urlmap.rb:58:in `call'
unicorn (6.0.0) lib/unicorn/http_server.rb:634:in `process_client'
unicorn (6.0.0) lib/unicorn/http_server.rb:732:in `worker_loop'
unicorn (6.0.0) lib/unicorn/http_server.rb:547:in `spawn_missing_workers'
unicorn (6.0.0) lib/unicorn/http_server.rb:143:in `start'
unicorn (6.0.0) bin/unicorn:128:in `<top (required)>'
vendor/bundle/ruby/2.7.0/bin/unicorn:23:in `load'
vendor/bundle/ruby/2.7.0/bin/unicorn:23:in `<main>'

The following line seems to be the culprit. But since the plugin is not enabled, it should not fire, right?

When discourse-prometheus-alert-receiver is enabled, the site works fine. When it is not enabled, the site breaks. This sounds like a bug to me.