Date: Mon, 8 Nov 2010 16:57:09 +0300 From: cronfy <cronfy@gmail.com> To: freebsd-hackers@freebsd.org Subject: Re: Slow disk access while rsync - what should I tune? Message-ID: <AANLkTin2z_hXrjd1OsrWZAm7Yi4sOK4Ak_=ZLq1wBSTM@mail.gmail.com> In-Reply-To: <201010302237.o9UMb157032371@apollo.backplane.com> References: <AANLkTikzZvZn=vNNRtcSViWq8ty7b8qOooQ4NbHiJH5q@mail.gmail.com> <AANLkTikpk4O-q_Omh9eAGZB474J1BVu2YJ7OKWvhZm7v@mail.gmail.com> <201010302237.o9UMb157032371@apollo.backplane.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hello. Thanks again. > Yes, hardlinked backups pretty much destroy performance, mainly > because it destroys all locality of reference on the storage media > when files are slowly modified and get their own copies, mixed with > other 'old' files which have not been modified. But theoretically > that should only effect the backup target storage and not the server's > production storage. That is what surprised me when I did experiment with backups. If I move backup off from the production server (to another less loaded production server indeed), server that shuld be backed up starts to run fine while backups are created. I think it means that problem is not with vnode/dir caches.. At the other side the server who received backups became very slow. So the problem looks to be related to writes or file creation/hardlinking somehow... At the moment I do not have server with ZFS, but I will think in this direction. But I heard that ZFS has less performance than UFS, is it really like this? I mean I have seen benchmarks and system requirements, but would like to hear about your own experience. -- // cronfy
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AANLkTin2z_hXrjd1OsrWZAm7Yi4sOK4Ak_=ZLq1wBSTM>