Date: Sun, 7 Jun 2009 08:53:19 +0200 (CEST) From: Wojciech Puchar <wojtek@wojtek.tensor.gdynia.pl> To: Sergio de Almeida Lenzi <lenzi.sergio@gmail.com> Cc: freebsd-questions <questions@freebsd.org> Subject: Re: fsck on 1.5TB drive Message-ID: <alpine.BSF.2.00.0906070848390.97754@wojtek.tensor.gdynia.pl> In-Reply-To: <1244340648.8936.7.camel@localhost> References: <56942.76.25.231.251.1244295367.squirrel@webmail.wcubed.net> <1244306366.5333.8.camel@localhost> <alpine.BSF.2.00.0906062312520.93574@wojtek.tensor.gdynia.pl> <1244340648.8936.7.camel@localhost>
next in thread | previous in thread | raw e-mail | index | archive | help
> > > You are right Puchar, but sometimes (2 in 100 on powerfailure) the > filesystem > gets corrupted (database files opened, and being extended)... so > when the fsck enters, the database get corrupted.. Filesystem will rather be not corrupted, but database file data. Non-journalled UFS with softupdates guarrantes the right sequence of disk updates. For example it will not allocate just freed space until freeid inodes/blocks are not wrote back to disk. As in your example - extended and written something, but will end unextended etc.. > by using zfs or journaling I never have anothter database problem.... This is sequence problem - for example you write to file A,B and C then it's a crash and you have file A and C written but not B. I though that all this "famous" database systems like mysql already have mechanism for that. looks like not, or it should not get corrupted.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.0906070848390.97754>