Date: Sat, 13 Jul 2002 22:04:18 +0100 From: Matthew Seaman <m.seaman@infracaninophile.co.uk> To: nathan@vidican.com Cc: questions@FreeBSD.ORG Subject: Re: Idea: using CVSUP to mirror websites accross multiple servers... opinions? Message-ID: <20020713210418.GB22908@happy-idiot-talk.infracaninophi> In-Reply-To: <200207132023.g6DKN0A23446@mail.ipsnetwork.net> References: <200207132023.g6DKN0A23446@mail.ipsnetwork.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jul 13, 2002 at 04:23:00PM -0400, owner-freebsd-questions@FreeBSD.ORG wrote: > I was just thinking of a way to effectively mirror content accross > multiple webservers in a load-sharing environment. > I was figuring on keeping a single FTP server for the master copy > and then replicating the data accross multiple machines. I'd rather > copy than use NFS; so there would not be a single point of failure, > (eg: a drive on the ftp server croakes). I was thinking of using > rsync; but from the way I understand it rsync just re-downloads the > entire tree? Could I not use CVSUP to accomplish the replication of > the data? As-in un a CVS server from the FTP server machine, and have > the webservers sync the trees they need? Any issues with using CVS and > image files? Or binaries alltogether? Any ideas or suggestions for a > better method for data replication and synchronization? You can certainly use rsync to keep your servers in sync, and it will use the rsync algorithm (http://rsync.samba.org/tech_report/) to minimise the amount of data actually copied over the wire. That means that files which are already present on the server won't be unnecessarily copied, even that only those chunks of a large file that actually contain changes will be copied. Works very well indeed, especially if you run it over ssh --- see the FAQ at http://www.snailbook.com/faq/no-passphrase.auto.html for details of how to secure unattended logins. You can use cvsup in exactly the same way, but unless you've got all your content saved into a CVS repository, cvsup will pretty much fall back to running just like rsync. If you are using CVS, cvsup will look at the $Revision$ tags and pull out precisely the diff to the latest version and apply it, which is even more efficient than rsync. Binary file types --- images and so forth, which you can't put CVS tags into --- can be used with a cvs repository. Remember to check-in binary files using the `-kb' option. Binary files will always cause cvsup to fall back to the rsync algorithm. Other alternatives? None that I can think of better than rsync or cvsup. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. 26 The Paddocks Savill Way Tel: +44 1628 476614 Marlow Fax: +44 0870 0522645 Bucks., SL7 1TH UK To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020713210418.GB22908>