Description
Right now is_tuple_expired (is_expired in cartridge role configuration) is a required argument when configuring an expirationd worker
It must be a name of a function in _G.
Right now the problem is that there is at all a need to code such functions for the developers of an app.
What if you don't wan't to develop an app, but rather want to deploy a default cartridge cluster (just straight up create -> pack -> deploy) that has everything else you need provided via other modules, and configure it? In other words, a full out-of-the-box experience.
A configuration like this:
ddl/migrationsfor applying schemacrudfor data accessexpirationd(??? need to write own role that goes together with expirer OR write code ininit.lua)
How about implementing a bunch of such functions that are provided by defalut?
Something like
local M = {}
local fiber = require('fiber')
function M.check_ttl(args, tuple)
local now = fiber.time()
local ttl = tuple[args.ttl_field]
local timestamp = tuple[args.timestamp_field]
if (ttl == nil or timestamp == nil) then
return false
end
return now > timestamp + ttl
end
function M.check_time_to_remove(args, tuple)
-- same thing, but for tuples
-- that only have a "time to expire" timestamp, instead of the "creation time + ttl" combo
end
return M
...
-- somewhere in expirationd
for f_name, f in pairs(require('expirationd.functions') do
rawset(_G, '__'..f_name, f)
endSo the possible expirationd.yml configuration would look like:
# tasks
expire_at:
space: session
is_expired: __check_ttl
options:
args:
ttl_field: max_inactive_interval
timestamp_field: last_accessed_time
is_master_only: trueWith no need to code anything in cartridge init.lua or in a custom role