* Mon, Sep 08, 2003 - 10:47:07 -0500, Dave Kleikamp: > I think I know what's going on. The space isn't freed until the journal > record for the remove transaction has been committed to disk. To reduce > I/O to the journal, the journal is often not written to until a journal > page is completely filled. It's often the case that other activity in > the fs causes the journal to be written, but not in your case. >
> I think I need to have something trigger a write of the journal if there > is uncommitted data there and a small amount of time has passed. I actually found out by writing a 700MB file to a file system with 1GB of size, removing it and again copying that file into that file system, where copy then aborted with only 300 MB written. But one could do that on small file systems (e.g. /boot) very fast and just triggering by time could fail then too. Erwin _______________________________________________ Jfs-discussion mailing list [EMAIL PROTECTED] http://www-124.ibm.com/developerworks/oss/mailman/listinfo/jfs-discussion
