On Sat, Jun 7, 2008 at 8:22 PM, Corin Moss <[EMAIL PROTECTED]> wrote: > > As far as general feedback regarding ruote-rest is concerned - it's > great! In my prototype I'm actually clustering multiple instances of > ruote-rest behind a load balancer, so of course it being HTTP makes > things a bit easier - I have to admit I've not looked into the native > clustering / balancing capabilities I've seen mentioned in Openwferu-users.
Neat. In fact there are no "native" clustering / balancing capabilities, I think what you've seen in the list were discussions and recommendations. I'm just trying to stay out of the way of people doing clustering / load-balancing, same thing goes for authentication and authorization. BTW, what is your approach to that ? If you have some time to share your experience with your load balancing solution, that would be very appreciated. > I've paired it with rest_client (http://rest-client.heroku.com/rdoc/) > which cites Sinatra as inspiration - and it all works perfectly so far! There's also rufus-verbs : http://rufus.rubyforge.org/rufus-verbs (shameless self-promotion) but no worries, rest-client is neat and the Heroku stuff is great, I'm very happy you can leverage ruote-rest with rest-client. > The example you gave for wait_for is indeed what I was thinking about. > I understand the warning regarding early termination - I'm guessing this > is when a flow is executed so quickly as to finish before wait_for is > called? exactly. > I wonder if an alternative would be to define a method on the engine - > something like launch_and_wait_for? I guess one of the big problems > with this approach is it removes any of the minute control of the > process that one would normally have. Your idea got me thinking about optimizing wait_for to make it snappier. I think that the current implementation can be enhanced and thus avoid those "process terminates before call to wait_for" moments. I can't make the engine synchronous, it would make the ruby call stack reach its ceiling very fast (in fact, the first versions of ruote were very naive and the ceiling was reached, happily I could switch the implementation to a workqueue, and now it's as it should be). I will try to enhance wait_for and notify you if I'm successful. http://rubyforge.org/tracker/index.php?func=detail&aid=20554&group_id=2609&atid=10195 For now I guess that it already works fine most of the time (though I don't know what your 'syncrhonous' processes look like). You haven't talked about the "result" of the synchronous process. Do you store that result somewhere ? Do you want to include it in the reply to the POST ? Cheers, -- John Mettraux - http://jmettraux.wordpress.com --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "OpenWFEru users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/openwferu-users?hl=en -~----------~----~----~----~------~----~------~--~---
