Multiple delayed jobs using in a ruby class

[amtest]

I have one sudo code here, using delayedjob with rails, actually tired but in some cases it getting weird, i want to call the first delay and after that is finished, then only start second delay, so i have used the priority. Is this the proper way to do this or is anything wrong? class Core def foo priority = 0 if condition1 if condition2 self.delay(priority: priority).foo1(priority) end self.delay(priority: priority).foo1(priority) end end def foo1(priority) if condition1 if condition2 priority +=1 self.delay(priority: priority).foo1(priority) end priority +=1 self.delay(priority: priority).foo1(priority) end end end

[Paul Makepeace]

If you want to sequence the jobs why not just make it one job where the second "job" is just a method call? One important purpose of asynchronous jobs like this is to get them out of the way of a main loop. By making it one job you've done that so there's no obvious (so far) pressure to punt the other one since you're not in that main loop any more.

Your pseudo code leaves me with more questions than answers so maybe explain what you're trying to actually do?

--
You received this message because you are subscribed to the Google Groups "Ruby on Rails: Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rubyonrails-ta...@googlegroups.com.
To post to this group, send email to rubyonra...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rubyonrails-talk/76889966-18f5-40b7-a112-036a18679f83%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[amtest]

Hi, thanks for the response paul, actually i am trying to do export the csv, so that will take bit time to getting data and push to s3 that's why i put the delay, so one delay is fine to doing this long job? is this the normal way basically i am get rid of this multiple delay, because some times it wrap up wrong data

[Paul Makepeace]

You could try something like,

   def export_csv_and_push_to_s3

      export_to_csv unless csv_exists?

      push_to_s3

The fact it takes a long time shouldn't matter (so long as you've set your DJ timeout appropriately). Presumably you're storing your CSV somewhere locally and your object has a way of retrieving that (csv_exists?) If any part of export_csv_and_push_to_s3 fails it'll skip the CSV generation next time if it has completed it already.

You could optionally delay push_to_s3 altho I'm not really seeing the benefit.

Either way you definitely should not (IMO) be attempting to orchestrate some sequence during the original job scheduling.

To view this discussion on the web visit https://groups.google.com/d/msgid/rubyonrails-talk/90f4fdb9-7ea8-4232-af4a-5dfdf8bd9840%40googlegroups.com.