Hi Guys, I've built a client-side Distributed storage system with 2 servers and I've ran into a problem. ServerA has about 3GB of free space, and ServerB has 600mb. I'm trying to copy a 1GB file into the storage system, but it keeps trying to copy it into ServerB. How do I go about fixing this? What is the best distribute set up for maximum scalability and reliability? Regards, Simon -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20090326/50744326/attachment.html>
2009/3/26 Simon <simonl at bigair.net.au>:> Hi Guys, > > > > I?ve built a client-side Distributed storage system with 2 servers and I?ve > ran into a problem. > > ServerA has about 3GB of free space, and ServerB has 600mb. > > > > I?m trying to copy a 1GB file into the storage system, but it keeps trying > to copy it into ServerB.Are you using distribute, replicate, or distribute+replicate? If you are using replicate then it cannot obviously create a 1Gb file if only 600mb is available. Vikas -- Engineer - Z Research http://gluster.com/
2009/3/27 Simon <simonl at bigair.net.au>:> Hi Vikas, > > I am currently just testing a basic distribute with 2 servers.Distribute schedules files to servers based on a hash of the inode number. In your case it appears that the file has been scheduled to server B by the hash algorithm. Vikas -- Engineer - Z Research http://gluster.com/