Fork me on GitHub
#aws
<
2019-05-02
>
hiredman00:05:34

Step functions can wait till a certain time or until an amount of time elapses then call a lambda

wei03:05:43

will check out AWS step functions, thanks.

lloydshark07:05:28

@wei AWS Lambda can be triggered by CloudWatch events (could easily add extra logic inside the lambda if needed).

lloydshark07:05:28

We used Lambda via CloudWatch for any time based triggering at my last work.

wei07:05:34

@lloydshark thanks, I'm concerned about the 100 rule limit on cloudwatch events though, seems better suited for recurring events. for my use case I'd need to schedule thousands of one-time events

valtteri07:05:04

What is the timespan? I mean, how long time there’s between receiving the event and handling it?

wei08:05:59

could be weeks or months

valtteri11:05:46

In that case probably the simplest solution is to use Step Functions as hiredman mentioned. If the timespan was less than 30 days you could’ve abused SQS as well. DynamoDB TTL feature sounds like an interesting alternative though! Is the idea that you use delete events in the DDB Stream as triggers?

valtteri07:05:14

(in case it’s known)

valtteri07:05:58

And how precise does the triggering need to be?

lloydshark07:05:58

If it really suits you can ask for an increase in the 100 rule limit.

wei08:05:30

I've applied so we'll see what they say, not sure if I'll be able to get an order of magnitude more though

lloydshark10:05:44

Yeah I don't know what the hard limit is...

lloydshark07:05:41

But like @valtteri says - there might be other options depending on what you actually need to do.

viesti13:05:04

while googling around, ran into aws simple workflow, which has eerily ~similar limits https://docs.aws.amazon.com/amazonswf/latest/developerguide/swf-dg-limits.html

viesti13:05:07

don't know how the cost of quartz + ec2 compare to these, probably depends on the number of tasks to be scheduled