Date: Sat, 2 Jun 2007 23:38:50 -0500 From: Craig Boston <craig@yekse.gank.org> To: Thomas Hummel <hummel@pasteur.fr> Cc: ports@freebsd.org, Paul Schmehl <pauls@utdallas.edu> Subject: Re: Uggg! Message-ID: <20070603043850.GB1341@nowhere> In-Reply-To: <20070601154057.GY43953@parmesan.sis.pasteur.fr> References: <200706010521.l515LE4N074880@harmony.bsdimp.com> <20070601085750.ang0g5aqp0kg8c8k@webmail.leidinger.net> <20070601083345.GA48323@rot13.obsecurity.org> <10723ADA-FD53-45F8-BDFA-DBD98CBC212E@FreeBSD.org> <465FFE06.7010900@gmx.de> <20070601132205.GW43953@parmesan.sis.pasteur.fr> <6D9E1847C218B06AE67CB299@utd59514.utdallas.edu> <20070601154057.GY43953@parmesan.sis.pasteur.fr>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Jun 01, 2007 at 05:40:57PM +0200, Thomas Hummel wrote: > Sure. But that doesn't explain why so many +CONTENT files were screwed > up and why there isn't a easy or easier way to re-generate them. Sound like pretty standard behavior of softupdates to me. Often files that were created within the last (memory fuzzy) seconds will be lost. I suspect in both cases the battery died right after an upgrade finished and portupgrade tweaked the +CONTENTS files of pretty much everything in order to update dependencies. I'd wager it's doing a create temp file/delete/rename dance. As far as I can tell, having files disappear is considered better than ending up with a corrupt filesystem, so sayeth the FS gods. It might be interesting to see how UFS+gjournal/ZFS stack up under similar circumstances. Craig (apologies if anyone has posted this already, I'm only about 1/3 down the thread)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070603043850.GB1341>