When debug mirror process, queue creation failure with resource temporarily unavailable

[lunny]

2022/02/26 20:34:50 .../cron/tasks_basic.go:41:func1() [I] ======== &cron.UpdateMirrorTaskConfig{BaseConfig:cron.BaseConfig{Enabled:true, RunAtStart:false, Schedule:"@every 1m", NoSuccessNotice:true}, PullLimit:1000, PushLimit:1000}
2022/02/26 20:34:50 ...ces/mirror/mirror.go:56:Update() [T] Doing: Update
2022/02/26 20:34:50 ...odels/repo/mirror.go:128:MirrorsIterate() [I] [SQL] SELECT "id", "repo_id", "interval", "enable_prune", "updated_unix", "next_update_unix", "lfs_enabled", "lfs_endpoint" FROM "mirror" WHERE (next_update_unix<=$1) AND (next_update_unix!=0) ORDER BY updated_unix ASC [1645878890] - 31.802881ms
2022/02/26 20:34:50 ...s/repo/pushmirror.go:109:PushMirrorsIterate() [I] [SQL] SELECT "id", "repo_id", "remote_name", "interval", "created_unix", "last_update", "last_error" FROM "push_mirror" WHERE (last_update + ("interval" / $1) <= $2) AND ("interval" != 0) ORDER BY last_update ASC [1s 1645878890] - 15.163638ms
2022/02/26 20:34:50 ...ces/mirror/mirror.go:137:Update() [T] Finished: Update: 0 pull mirrors and 0 push mirrors queued
2022/02/26 20:35:09 ...ueue_disk_channel.go:165:Run() [F] Unable to create internal queue for notification-service Error: unable to create queue level for notification-service with cfg queue.LevelQueueConfiguration{ByteFIFOQueueConfiguration:queue.ByteFIFOQueueConfiguration{WorkerPoolConfiguration:queue.WorkerPoolConfiguration{QueueLength:20, BatchLength:20, BlockTimeout:1000000000, BoostTimeout:300000000000, BoostWorkers:1, MaxWorkers:5}, Workers:0, Name:"notification-service-level", WaitOnEmpty:false}, DataDir:"/data/gitea/gitea/data/queues/common", ConnectionString:"", QueueName:""} by max attempts: error: resource temporarily unavailable

[lunny]

OK. I see, different type queues should not share a folder.

[zeripath]

They should be able to share the same folder - common queues should be detected and just use the same connection. I'm gonna need to see the rest of your queue configuration.

This error looks like there's something else running at the same on the same dir.

[lunny]

duplicated with #18917

[zeripath]

I think in this case the problem was that there were conflicting CONN_STRs for the connection to the level db. Removing the incorrect CONN_STR from the queue configuration solves the issue but #18923 would prevent the problem too.