Hannu Krosing <hannu at skype.net> writes:
> ??hel kenal p??eval, E, 2006-02-13 kell 08:34, kirjutas Christopher
> Browne:
>> Hannu Krosing <hannu at skype.net> writes:
>> > ??hel kenal p????eval, E, 2006-02-13 kell 11:36, kirjutas Kevin Gill:
>> >> Is there a method for replicating the database users (pg_users)?
>> >
>> > pg_user is a VIEW and as such you will probably not be able to
>> > replicate it.
>> >
>> > You could try to replicate the underlying table (pg_shadow), though
>> > I'm not sure if slony allows you to do it.
>> 
>> The trouble with trying to replicate system tables is that you can't
>> put triggers on them, which is what Slony-I requires...
>
> perhaps it can be done by having a user table mirror pg_shadow, via some
> regular script, and then replicate that.
>
> then the failover process would need to do "begin; delete from
> pg_shadow; insert into pg_shadow select * from
> replicated_pg_shadow;commit;" to have a fresh copy

Yes, I guess that would work.

A way to do this that would be, oh, call it "forcibly robust" would be
for each SYNC to check to see if pg_shadow had changed, and update the
user table correspondingly.

In practice, it's probably good enough to have some script that runs
every so often to look for changes.  I know we don't modify pg_shadow
on our systems so often as every week; polling for it once a day would
likely suffice, for us...
-- 
"cbbrowne","@","ca.afilias.info"
<http://dba2.int.libertyrms.com/>
Christopher Browne
(416) 673-4124 (land)