Dear forum
I have a client who have a Servoy 5.2 / Sybase 11 solution which has been in production since Servoy 2.2. In November 2010 It was updated from 3.5 to 5.2, and from sybase 9 to sybase 11.
The server currently runs on an Intel XServe which has a virtualised installation of OS X Server purely for Servoy. The server works fine, but sometimes runs a little low on ram because another VM runs a Blackberry server on the same hardware.
About 3 or 4 times now, they have complained that Servoy will suddenly stop working and disconnect all (40) users. This has only happened since we moved to 5.2, and onto the VM, so it is not clear whether it is either that causes it.
Most recently, yesterday at 6pm, I imported a few VERY minor changes into the solution (added one text dataprovider, and moved 3 fields around on a form). The import began, but then after about 10 minutes, it failed (connection refused in the solution import window). We checked and apparently Sybase had unexpectedly quit on the server (it seems my import had caused it to crash)
I then retried the import and it hung in the solution import window after a few info lines. This continued for 1.5 hours. We left it thinking that it would eventually finish, and not wanting to disturb it if it was going to eventually work… The server its-self was swinging between 0% cpu and 140%cpu (and sometimes hanging around 36%) for dbsrv11. My network on my machine was idle so it wasn’t sending the file (which was only 1.6Mb). Obviously Sybase was very busy, and apparently it was recovering from the original crash on the first import (although it appeared that it had simply hung during my import)
After talking to Jan Blok about it, we eventually quit the browser window (after 1.5 hours), sybase eventually stabilised, and we were again able to open the solution, but without my changes (so the import was unsuccessful.)
The process though has doubled the size of one of the sybase .db files. It started out as 44GB (it is used for high res streamed images). The file is now 79GB. If these were small files, we wouldn’t bother too much, but being this size, and backing up offsite daily, this change will have an impact on backing up, and it is unnerving to know that there is that much data and we don’t know what it is.
It would be useful to know what anyone thinks that the database has actually done, and whether it might clean its-self up again, or will I have to unload / reload the database to fix this?
Also, dbsrv11 unexpectedly quitting is a significant issue. I have the sybase logs from the time when the server crashed yesterday, but don’t know what to look for.
Would be grateful for any advice on this…
Thanks
Bevil