On Fri, 2005-08-26 at 10:06 -0500, John Goerzen wrote: > On Fri, Aug 26, 2005 at 09:58:14AM -0500, Dave Kleikamp wrote: > > On Fri, 2005-08-26 at 09:51 -0500, John Goerzen wrote: > > > On Fri, Aug 26, 2005 at 09:30:33AM -0500, Dave Kleikamp wrote: > > > > On Fri, 2005-08-26 at 09:24 -0500, John Goerzen wrote: > > > > > > > > An xtree is the data structure within jfs's inode, which defines the > > > > data extents used to store the file data. It is possible that the xtree > > > > of an infrequently accessed file may have been corrupted earlier and > > > > gone undetected for a while. > > > > > > If this is correct, would you expect a fsck -f with the 1.1.8 tools to > > > detect and correct it? > > > > Yes, I would expect even 1.1.7 to detect & correct it. If you've run > > fsck -f with 1.1.7, I wouldn't expect 1.1.8 to find anything new. > > That is what I did today, and also what I did back when the server was > upgraded to 2.6. Hmm. But perhaps if 1.1.7 introduced the problem, it > wouldn't detect it on the same pass?
If your system crashed, and fsck was run without the -f, it could have introduced the problem (if it was caused by the log-replaying bug). The next time fsck -f was run (or the same run, if -f was specified initially), it should have found & fixed the problem. -- David Kleikamp IBM Linux Technology Center ------------------------------------------------------- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf _______________________________________________ Jfs-discussion mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/jfs-discussion
