cron

cron is a tiny server-side scheduler shipped with ESX Legacy. Other resources register a callback for a time of day (hour and minute) and cron calls it once every day at that time. Official addons use it, for example esx_society for salaries and esx_vehicleshop for rental payments.

OverviewCopy Anchor Link

cron:runAtCopy Anchor Link

The only API is a server-side local event.

TriggerEvent("cron:runAt", h, m, cb)

Callback arguments:

Wrong argument types fail with an assert that names the invoking resource.

ExamplesCopy Anchor Link

-- Every day at 05:10
TriggerEvent("cron:runAt", 5, 10, function(d, h, m)
    print(("Daily cleanup at %02d:%02d"):format(h, m))
    MySQL.update("DELETE FROM my_logs WHERE created_at < NOW() - INTERVAL 30 DAY")
end)

-- Every Monday at 18:30
local MONDAY <const> = 2

TriggerEvent("cron:runAt", 18, 30, function(d)
    if d ~= MONDAY then return end
    print("Weekly payout")
end)

-- Every 6 hours: register several times
for _, hour in ipairs({ 0, 6, 12, 18 }) do
    TriggerEvent("cron:runAt", hour, 0, function()
        print("Six-hour task")
    end)
end

How scheduling worksCopy Anchor Link

• cron checks once per minute, aligned to the start of the minute.
• On each new minute it computes every job's most recent scheduled time (today, or yesterday if today's time is still ahead) and runs the job if that time is newer than the previous check. A short server hitch therefore does not skip a job.
• Jobs do not catch up after a restart. A task scheduled for 05:10 does not run if the server was offline at 05:10.
• Errors in a callback are caught and printed as cron job at HH:MM errored, skipping it; other jobs keep running.

Common issuesCopy Anchor Link

• Weekday check never matches. The official README uses d == 1 for Monday. In Lua wday 1 is Sunday; Monday is 2.
• Task runs at the wrong hour. The server's timezone differs from yours. Check os.date() in the server console.
• Errors after restarting a resource. Jobs cannot be removed. When you restart a resource that registered a job, the old callback reference stays registered and errors every day, next to the new job the restarted resource registers. The error is harmless, but a full server restart cleans it up.
• Jobs stop after restarting cron. Jobs live in cron's memory. Restarting cron alone drops every job; the resources that registered them must be restarted too.
• Registering from the client. cron:runAt is a server-only local event; it does nothing on the client and cannot be triggered over the network.

RelatedCopy Anchor Link

• Core resources overview
• Server functions
• esx_society
• oxmysql guide

Frequently asked questions