On Thu, Sep 09, 2010 at 02:51:59PM +0200, Asier wrote: > El 09/09/2010 14:46, Gonzalo Suarez escribió: > > > >We haven't used ruote-kit or ruote 2.1.x much yet. But as far as I > >understand, the issue has little to do with ruote or ruote-kit. The > >problem appears when running ruote inside a Java application server > >with jruby. As Asier states, ruote-kit may be warbled and thrown into > >a Glassfish. But, what happens with the engine when I stop the > >Glassfish server? I guess two things should happen: > > > >1. The server stops listening > >2. The engine (ruote) should stop (persist all expressions to disk, > >close files, etc...) > > > >I think the second step would not happen. But this is a guess... I > >haven't tried it. I know this behavior happened to us with ruote-rest > >+ glassfish in the past... > > Uff, that could have dramatic consequences if a critical transaction is > broken :-o > > (and Murphy's law allways is against us) > > I don't know how to solve this.
>From the ruote point of view, a half broken workflow is OK, if it resulted in >a process error you can replay at it, if the process is simply "stalled", you >can re_apply the stuck segment. http://ruote.rubyforge.org/process_administration.html Now, if one of your participant is in a "transaction", well it all depends on how you implemented it. Participants should have a sensible "cancel(fei, flavour)" implementation... So not too dramatic if you're prepared for it. Cheers, -- John Mettraux - http://jmettraux.wordpress.com -- you received this message because you are subscribed to the "ruote users" group. to post : send email to [email protected] to unsubscribe : send email to [email protected] more options : http://groups.google.com/group/openwferu-users?hl=en
