On Wed, 2003-11-12 at 21:41, Joe Marcus Clarke wrote:
[jakobbg dune jakobbg]$ strace evolution 1>stdout 2>stderr File is available here: http://www.grimstveit.no/~jakob/files/temp/evol_straceI looked at the truss output, and it looks like Evo is still trying to read your "cur" mailbox. It could be that it's just choking under the size. Some users have reported that converting to maildir format helps alleviate such problems.
Thanks for the advice. * portupgrade -Rrf evolution ... did not fix this. However - more investigation on this topic lead me to find the problem: It's the indexing of a specific folder that goes into a loop, running till the device is full. Found this by using `du -hs *` in maildir hierarchy, locating where in the tree the disk consumption was occuring. After 20 minutes Evo had consumed 2Gb in a single temporary indexing file for that maildir (I have only maildirs, using indexing). No space left made indexing stop and Evo continued to load and start. The temporary indexing file was automatically removed, going back to 49% usage of /home. When going through all my local-metadata.xml's in each folder, changing <folder type="maildir" name="mbox" index="1"/> to «index="0"» instead, Evolution fired up without any problems. Disk space of ~/evolution/local went from 2Gb to 445Mb as well after this (very strange, probably some other stray indexing file lying around somewhere). Thus: Not actually a FreeBSD/Gnome issue, will report this erratic behavior to Ximian Evolution list, hopefully helping them to thwart a bug or two. Console output: http://www.grimstveit.no/~jakob/files/temp/evol_stdout Thanks for the help anyways :-) -- Jakob Breivik Grimstveit, http://www.grimstveit.no/~jakob, +47 48298152
Attachment:
signature.asc
Description: This is a digitally signed message part