starred/asynq

Fork 0

mirror of https://github.com/hibiken/asynq.git synced 2026-04-26 15:35:55 +03:00

[GH-ISSUE #923] [Question] Are there any way to run the tasks that's retention exceeds #1473

New issue

Open

opened 2026-03-07 22:09:49 +03:00 by kerem · 3 comments

kerem commented

2026-03-07 22:09:49 +03:00

Owner

Originally created by @arifmahmudrana on GitHub (Oct 23, 2024).
Original GitHub issue: https://github.com/hibiken/asynq/issues/923

Are there any way to run the tasks that's retention exceeds? e.g. when we call GetTaskInfo in the asynq.Inspector it returns the TaskInfo but when executing the task using the RunTask using the asynq.Inspector we get error. You are calling a redis script

var runTaskCmd = redis.NewScript(`
if redis.call("EXISTS", KEYS[1]) == 0 then
	return 0
end
local state, group = unpack(redis.call("HMGET", KEYS[1], "state", "group"))
if state == "active" then
	return -1
elseif state == "pending" then
	return -2
elseif state == "aggregating" then
	local n = redis.call("ZREM", ARGV[3] .. group, ARGV[1])
	if n == 0 then
		return redis.error_reply("internal error: task id not found in zset " .. tostring(ARGV[3] .. group))
	end
	if redis.call("ZCARD", ARGV[3] .. group) == 0 then
		redis.call("SREM", KEYS[3], group)
	end
else
	local n = redis.call("ZREM", ARGV[2] .. state, ARGV[1])
	if n == 0 then
		return redis.error_reply("internal error: task id not found in zset " .. tostring(ARGV[2] .. state))
	end
end
redis.call("LPUSH", KEYS[2], ARGV[1])
redis.call("HSET", KEYS[1], "state", "pending")
return 1
`)

So I assume it's erroring while calling local n = redis.call("ZREM", ARGV[2] .. state, ARGV[1]) because my task isn't in pending or active or in aggregating.

So are there anything I can do to add that to the queue may be by adding a record in the redis any idea?

Originally created by @arifmahmudrana on GitHub (Oct 23, 2024). Original GitHub issue: https://github.com/hibiken/asynq/issues/923 Are there any way to run the tasks that's retention exceeds? e.g. when we call `GetTaskInfo` in the `asynq.Inspector` it returns the TaskInfo but when executing the task using the `RunTask` using the `asynq.Inspector` we get error. You are calling a redis script ```go var runTaskCmd = redis.NewScript(` if redis.call("EXISTS", KEYS[1]) == 0 then return 0 end local state, group = unpack(redis.call("HMGET", KEYS[1], "state", "group")) if state == "active" then return -1 elseif state == "pending" then return -2 elseif state == "aggregating" then local n = redis.call("ZREM", ARGV[3] .. group, ARGV[1]) if n == 0 then return redis.error_reply("internal error: task id not found in zset " .. tostring(ARGV[3] .. group)) end if redis.call("ZCARD", ARGV[3] .. group) == 0 then redis.call("SREM", KEYS[3], group) end else local n = redis.call("ZREM", ARGV[2] .. state, ARGV[1]) if n == 0 then return redis.error_reply("internal error: task id not found in zset " .. tostring(ARGV[2] .. state)) end end redis.call("LPUSH", KEYS[2], ARGV[1]) redis.call("HSET", KEYS[1], "state", "pending") return 1 `) ``` So I assume it's erroring while calling `local n = redis.call("ZREM", ARGV[2] .. state, ARGV[1])` because my task isn't in pending or active or in aggregating. So are there anything I can do to add that to the queue may be by adding a record in the redis any idea?

kerem commented

2026-03-07 22:09:50 +03:00

Author

Owner

@kamikazechaser commented on GitHub (Oct 25, 2024):

because my task isn't in pending or active or in aggregating.

Yes, it might be completed. If you want to run a completed task, would copying the payload and running it again suffice?

@kamikazechaser commented on GitHub (Oct 25, 2024): > because my task isn't in pending or active or in aggregating. Yes, it might be completed. If you want to run a completed task, would copying the payload and running it again suffice?

kerem commented

2026-03-07 22:09:50 +03:00

Author

Owner

@arifmahmudrana commented on GitHub (Oct 28, 2024):

because my task isn't in pending or active or in aggregating.

Yes, it might be completed. If you want to run a completed task, would copying the payload and running it again suffice?

@kamikazechaser thanks for your suggestion yes I can clone the task that's one option but my question is whether it's possible without cloning e.g. by adding some record in some hashset etc. I would prefer directly adding records in the redis without going one by one in the task list

@arifmahmudrana commented on GitHub (Oct 28, 2024): > > because my task isn't in pending or active or in aggregating. > > Yes, it might be completed. If you want to run a completed task, would copying the payload and running it again suffice? @kamikazechaser thanks for your suggestion yes I can clone the task that's one option but my question is whether it's possible without cloning e.g. by adding some record in some hashset etc. I would prefer directly adding records in the redis without going one by one in the task list

kerem commented

2026-03-07 22:09:50 +03:00

Author

Owner

@kamikazechaser commented on GitHub (Oct 29, 2024):

Well technically you could control the states within the hashset, but this is not documented. Maybe have a look at Inspector.RunAllArchivedTasks and see if you can modify it for your use case.

@kamikazechaser commented on GitHub (Oct 29, 2024): Well technically you could control the states within the hashset, but this is not documented. Maybe have a look at `Inspector.RunAllArchivedTasks` and see if you can modify it for your use case.

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

starred/asynq#1473

No description provided.

Rows
Columns