(Illustration by Gaich Muramatsu)
Hi, I know this is a FAQ, but one year ago, the homepage said the same and I won't believe it's still so. Currently we are running an AFS installion on nearly 30 machines with Linux Debian as root fs. But last week I picked off the server and all was down. This event brought on a discussion about availability and afs doesn't have any mechanisms for high availability. Does coda has an Achilles' heel like afs the volume server? How is the knowledge about the volume managed and saved? We plan to use coda on the whole faculty to provide user home dirs and data. This would be nearly 200 machines and 1000 users. Is coda ready for such a usage or would we run in trouble? I read in any documentation coda respects the unix permission of files. Is this correct? With afs we have the problem secure files in a public directory like /etc/shadow aren't save. Everyone can access them and we could restrict the access to them without inhibit access to the whole directory. How does coda match with kerberos? I read some articles in the archive which tell about problems. Are there still problems? Is there a port to True64? Or does Coda run on alpha architecture? Is it possible to deactivate the cache on the client machines at runtime? Does coda need a process in userspace to access the coda fs? Does coda support the new 64 bit system commands like fcntl64? Bye, Joerg. -- Wer einen Traum verwirklichen will, muss erst aufwachen.Received on 2003-11-29 16:57:06