Filemaker 12 conversion: Application crashes and windows scripting

Discussion created by BTimm on Mar 5, 2013
Latest reply on Mar 6, 2013 by BTimm

After much lurking the past year on all the 12 conversion issues and debates, I recently converted a deep, aged and somewhat complex solution from v11 to v12. I did a fair number of QA rounds testing various important layouts and processes and did not come across any huge show-stoppers.


I launched this past weekend and I'd say 98% of my solution is running fine and users seem to be working OK, although there are some legacy routines with start up and navigation that are clearly running a little but slower. Just slow enough to be noticeable, but not an immediate crisis.


More of an issue is some of the routines / scripts that are sporadically causing outright application freeze and need to force close. I'm perturbed by the volume of force closes I'm suddenly having to do, as well as the application crashes, which I haven't seen since we were on pre-v7. I've been able to narrow down one of the more critical problems with a specific script on one database. However, that script has a long tree of related, nested functions and scripts with things like creating portal rows, setting related fields, opening related records to run external scripts, etc.


It does not fail every time. I am able to run it successfully about 50% of the time, and suspect that it is something conditional with the types of records it's reviewing and need to narrow it down to specific script pothole.


Without going into details on this script, I was hoping some power users here might be able to give me updated pointers on specific areas to look for.


Windows navigation: I've looked at some the new 12 features for selecting windows, but expect that the conversion would put any select window step into document mode by default and allow smooth conversion of pre-12 windows script steps.


Server performance: we've also updated server 12 and are now running it in a Xen environment. My IT guy says we could possibly be dealing with an overloaded Xen server and need to throw more memory at it. I'll be looking at that, but is it possible that poor server memory would cause the sporadic crashes as side effect of processing delays?


I will post more here as I debug and get to some resolutions.