Date: Sat, 05 May 2001 05:52:51 -0700 From: Cy Schubert - ITSD Open Systems Group <Cy.Schubert@uumail.gov.bc.ca> To: stable <stable@FreeBSD.ORG> Subject: Re: soft update should be default Message-ID: <200105051253.f45CrHW01020@cwsys.cwsent.com> In-Reply-To: Your message of "Sat, 05 May 2001 00:58:39 PDT." <Pine.BSF.4.33.0105050055420.16321-100000@sdmail0.sd.bmarts.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <Pine.BSF.4.33.0105050055420.16321-100000@sdmail0.sd.bmarts.c om>, Go rdon Tetlow writes: > On Fri, 4 May 2001, Nick Sayer wrote: > > > Mike Smith wrote: > > > > >> Hi, > > >> > > >> Why 'soft update' is not default? > > >> It adds performance and stability, doesn't it? > > > > > > > > > It requires disabling of write caching, which typically reduces > > > performance (significantly). > > > > > Why wouldn't a similar requirement (disabling write caching) apply to > > non-softupdates filesystems? The disk doesn't know whether the write is > > synchronous or not, after all. > > That's the thing about it. If you have write-caching enabled then all bets > are off in any case (ie, you might as well run in async). In addition, > soft updates does some really smart things even over async. I sense (maybe incorrectly, so forgive me if I sense wrong) that people think that w/o softupdates turned on that FreeBSD performs sync writes. By default the *BSDs perform synchronous metadata (SMD) writes. That means that metadata, e.g. inode and directory writes are synchronously written, while the data itself is asynchronously written. Softupdates orders the metadata writes in such a manner that batches of metadata writes, each batch being synchronously written, are asynchronously written out. The data itself, e.g. user data, is asynchronously written out when SMD or softupdates is used. If you have a critical database, it is no more protected by softupdates than SMD. In that case a DBMS would need to issue fsync() after each commit or the filesystem would have to be mounted synchronously. Of course as Gordon writes above, all bets are off if your disk does write-caching. There is an excellent paper entitled, Soft Updates: A Solution to the Metadata Update Problem in File Systems, by Gergory R. Granger and Yale N. Patt at EECS, University of Michigan. The paper is at http://www.ece.cmu.edu/~ganger/papers/CSE-TR-254-95/. Regards, Phone: (250)387-8437 Cy Schubert Fax: (250)387-5766 Team Leader, Sun/Alpha Team Internet: Cy.Schubert@osg.gov.bc.ca Open Systems Group, ITSD, ISTA Province of BC To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200105051253.f45CrHW01020>