On Sunday 10 January 2010 16:00:39 Ron Aaron wrote: > Whatever I do, a 'fossil sync -R repo' does not copy over the user > passwords. So doing a 'failover' scenario where I have one repo backing up > another, is only correct as far as data goes, but users will be unable to > log-in.
So I found in the docs where it says not copying the user information is on purpose. And in the main, that's the right thing to do. Nevertheless, this precludes easily doing a 'fail-over' scenario. And I assume that in the Fossil self-hosting repos, there is some provision for copying user data from one database to another -- else how would one be able to log-in to whichever one chose? Obviously, one could do a SQL query to dump the user table from the 'main' server, and populate the "secondary" servers from it. But what if the main server is down, and a secondary is used to add a user? Or is that simply not a scenario Fossil is meant to deal with? One of the advantages I see to Fossil is the ability to make a repo robust by distributing it across (possibly geographically distant) servers. For the kinds of uses I have in mind, this is an important feature. So being able to propagate the metadata is something I would also like to see. Sorry for the long-winded mail... Best regards, Ron -- Sending me something private? Use my GPG public key: AD29415D
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ fossil-users mailing list [email protected] http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
