Travis Butler Posted April 5, 2010 Posted April 5, 2010 FM Server 10.0v2 running on Windows Server 2008 Enterprise SP2, Xeon x3220 2.4ghz, 8 GB RAM; Scriptmaster 3.32. Running into a problem where FM Server crashes when I execute a script on the server side that uses Post XML Data; when I run it client-side, it runs without apparent trouble. As far as I've been able to narrow it down with both server and script logging, the crash happens right after I call PostXMLData. The crash itself is just the server process, not the OS; I can open the FM Server Admin Console either remotely, or when RDP'ed into the server, and restart the server process without any trouble. The script uses PostXMLData to retrieve an updated category listing from eBay; I limit the call to 27 'parent' categories to cut down the amount of data retrieved, but the returned XML file still comes to about 526K when I run the script client-side. Is there a limit to how large a return file PostXMLData and/or Scriptmaster/Groovy can handle when running in the server environment, as opposed to client-side? Many thanks!
Smef Posted April 6, 2010 Posted April 6, 2010 I don't think that there is a difference in memory available for the server rather than the client. I would, however, recommend that you make sure you have the most updated version of Java on your server. Memory usage was changed recently in J6_18 and 19, so if you have an older version of Java it may be what is causing the difference between your client and server machines.
Travis Butler Posted April 7, 2010 Author Posted April 7, 2010 Unfortunately, updating Java on the server didn't seem to fix the problem; running the script server-side crashed the server just as it did before, as far as I could tell; the server log shows the identical sequence of events. My application's own internal logging doesn't show any events, but there aren't any script steps between the last event covered in the server log and my own logging step, so my guess is that the server crashed before the new record was written to disk. (Which is why I was sure to test this when no one else was using the system!) Any further idea what might be happening? I can email you a printout of the script if that'd help.
Travis Butler Posted April 19, 2010 Author Posted April 19, 2010 Just had a second script start crashing the server when it got a large reply back from eBay (this time close to 1 MB of text, getting a reply of 1389 auction notifications from GetSellerEvents). Again, script runs fine on the client, but kills the server process. Thoughts, feedback, ideas?
Smef Posted April 20, 2010 Posted April 20, 2010 You may want to try increasing the memory being used if this is happening on large replies. We have instructions for how to do this at http://wo.360works.com/cgi-bin/support/productsupport.cgi/Heapspace_Out_of_Memory_Error
Jesse Barnum Posted April 28, 2010 Posted April 28, 2010 Travis, is there a crash log available? This would be in /Library/Logs/CrashReports, I think.
Travis Butler Posted May 4, 2010 Author Posted May 4, 2010 Sever 10 Advanced is running on Windows Server 2008, not OS X Server, so it's not doing CrashReporter logs. The item our administrator found in Server 2008's Event Viewer that appears to be related to the crash is as follows: Log Name: Application Source: FMSocket_SL55 Date: 5/4/2010 5:00:56 PM Event ID: 0 Task Category: None Level: Error Keywords: Classic User: N/A Computer: BBEFMSERVER Description: The description for Event ID 0 from source FMSocket_SL55 cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer. If the event originated on another computer, the display information had to be saved with the event. The following information was included with the event: FMSocket_SL55@localhost,ErrorCode=3048,ErrorMessage=Internal error, failed to initialise the event trace component. Event Xml: 0 2 0 0x80000000000000 14492477 Application BBEFMSERVER FMSocket_SL55@localhost,ErrorCode=3048,ErrorMessage=Internal error, failed to initialise the event trace component. Does this help any?
Travis Butler Posted May 4, 2010 Author Posted May 4, 2010 (Oh, we did try the instructions in the Heapspace support document; but under Windows Server 2008 we weren't quite sure if we were putting it in the right folder, with the way Server 2008 aliases the older folder paths. In any event, what we tried didn't prevent the crash.)
Jesse Barnum Posted May 6, 2010 Posted May 6, 2010 Hi Travis - unfortunately, there is no useful information in that event log entry. I think the only way to troubleshoot this is going to be to try to reproduce the problem at our office and then run it with a debugger. Is it possible for you to create a small FileMaker file that just does the server-side process, verify that it still causes the problem, and then send it to us?
Travis Butler Posted May 6, 2010 Author Posted May 6, 2010 Here's the sample file; the script in question is "Pull Category XML List from eBay Test". While the script tries to do a debug log at key points in the script, in practice I've found logging results to be intermittent, depending on whether the server manages to write the log records to disk before the crash. Let me know if you have any further questions! eBay_Problem_Test_File.fp7.zip
Travis Butler Posted June 17, 2010 Author Posted June 17, 2010 Hi, Jesse, Any ideas yet on what's causing the crash?
Recommended Posts
This topic is 5291 days old. Please don't post here. Open a new topic instead.
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now