On Saturday 12 January 2002 20:53, Jeremy Zawodny wrote:
> On Fri, Jan 11, 2002 at 08:08:51AM -0800, Kyle Hayes wrote:
> > We write our own replication system using our own checkpointing and
> > code. We start with the update logs (we're not happy about the
> > future removal of the update logs b
On Fri, Jan 11, 2002 at 08:08:51AM -0800, Kyle Hayes wrote:
> We write our own replication system using our own checkpointing and
> code. We start with the update logs (we're not happy about the
> future removal of the update logs because of this). I've looked
> through the MySQL source to see
Hi!
Two-way replication is complex, and even more difficult is coping with
broken communications.
MySQL only supports one-way replication (master -> slaves). Auto-inc column
values are communicated in the binlog to the slaves.
Some ideas:
- Generate primary keys where you concatenate the id of
On Thursday 10 January 2002 17:22, Matthew Walker wrote:
> Where I work, we're at the point of needing to maintain two database
> servers, one inhouse, and the other at another physical location. We
> need to replicate between the servers, to keep them both up to date, as
> both will be getting up
Where I work, we're at the point of needing to maintain two database
servers, one inhouse, and the other at another physical location. We
need to replicate between the servers, to keep them both up to date, as
both will be getting updated. This leaves us with the problem of
resolving primary key c
