-
Bug
-
Resolution: Unresolved
-
Major
-
None
-
2.15.4 GA
-
None
-
False
-
-
False
-
Not Started
-
Not Started
-
Not Started
-
Not Started
-
Not Started
-
Not Started
-
-
-
Low
Sometimes after start-up Sidekiq pods produce errors in the log like:
ThinkingSphinx::QueryError, "unknown column: 'sphinx_internal_class_name' - REPLACE INTO account_core (id, `sphinx_internal_class_name`, `name`, `account_id`, `username`, `user_full_name`, `email`, `user_key`, `app_id`, `app_name`, `user_id`, `sphinx_internal_id`, `sphinx_internal_class`, `sphinx_deleted`, `sphinx_updated_at`, `provider_account_id`, `tenant_id`, `state`) VALUES (14978, 'Account', 'org_name-lkdjfggf-serch-yyith8w', '156', 'username-lkdjfggf-serch-xskw25c', ' ', 'username-lkdjfggf-serch-xskw25c@anything.invalid', 'af61253a5933dd9ba30bebb6661d16d1 ede2926ef59ecdc45da4a8960164ffc2', '3173be1c fc507eb9', 'org_name-lkdjfggf-serch-yyith8w's App ui_accou-lkdjfggf-accont-vmt8wba', '180', 156, 'Account', 0, 1697580778, 2, 2, 'approved')"
Restart of affected pod usually fixes the issue.
A potential cause for this is load order of thinking-sphinx indices vs the monkey patch we have for removing unnecessary `sphinx_internal_class_name` column from RT indices.
In 2.16 we upgrade to thinking-sphinx 5.5 which has our customizations built in.
Need to test that behavior is actually compatible to what we have . This means for QE that searching should work after upgrade for at least plans and services as a STI and non-STI models.
