It came to our attention during the retreat that a large number of people were not aware of the Jaquith tertiary archiver. This message is intended to get you started on the basics of Jaquith and encourage you all to take advantage of the resource. I. Why should I use Jaquith? 1) Jaquith provides access to massive storage from your own workstation. You simply put and get unix files from a mythical place called Jaquith from your own machine without worrying about the details. 2) Jaquith is extremely simple to use. You type "jput myfile" to archive a file, "jget myfile" to retrieve one, and "jls" to see what you have archived. There are more sophicated options but those three simple commands will do most of the trick. 3) Jaquith is robust. It has been in use for one year and there have been three instances of data loss, all of which occurred earlier on and were due to munged tapes. We will provide you with means of preventing that from happening. 4) Jaquith never deletes files. You can get to an arbitrary version of a file you have archived. 5) Jaquith implements protected access so files can be kept private. II. How do I sign up? Create an account for me on your machine (so I can install Jaquith for you). The passwd line is: rywang:EXjyru0mKUZag:5527:16:Randolph Wang,608-4 Evans,28284,525-2215:/home2/rywang:/bin/csh Send mail to rywang@cs.berkeley.edu. The message should contain: 1) A list of people who will be using Jaquith, 2) A list of machine addresses from which people will be initiating Jaquith requests, 3) Whether you want your data to be replicated (unreplicated data works faster but there is a slim chance of losing it). III. How do I use it? Jaquith is organized as a number of "archives", each of which is like a file system in the sense that it has its own "root". Each archive has its own directory structure. Users don't have to create or change Jaquith directory by hand (hence the lack of "jcd" or "jmkdir"). Instead, Jaquith automatically mirrors the unix directory from which Jaquith operations are performed. Suppose your current working directory is /home/me and you typed the following sequence of commands: jput foo jget foo jls Upon seeing the jput command, Jaquith looks for Jaquith directories /home and /home/me in the default archive. If they are not found, they are created. And finally Jaquith stores foo in the Jaqauith directory /home/me. Upon seeing the jget command, Jaquith intelligently retrieves /home/me/foo from the default archive and puts a copy of foo in the user's current working directory. Similarly, jls lists the contents of Jaquith directory /home/me.