Martin, I'm guessing you mean 1 database per table type.
On Tue, Feb 10, 2009 at 5:17 PM, Martin Gainty wrote:
>
> I vote for 1 table per TableType
> this will keep your DB schema consistent with Architecture
>
> Martin
> __
> Disclaimer and confidenti
I'll take that on board.
Thanks for your advice, mysql-master-master, Maatkit, mysqlperformanceblog,
your patches and community support!
On Tue, Feb 10, 2009 at 3:54 PM, Baron Schwartz wrote:
> Hi Michael,
>
> On Mon, Feb 9, 2009 at 6:03 PM, Michael Addyman
> wrote:
, Feb 10, 2009 at 2:57 PM, Michael Addyman <
> michael.addy...@googlemail.com> wrote:
>
>> Hooray! http://code.google.com/p/mysql-master-master/
>>
>> Am I crazy to be considering replicating 500+ databases? I think so...
>>
>
> I don't think the number
Hooray! http://code.google.com/p/mysql-master-master/
Am I crazy to be considering replicating 500+ databases? I think so...
On Tue, Feb 10, 2009 at 12:11 PM, Michael Addyman <
michael.addy...@googlemail.com> wrote:
> Walter, this is exactly why we went for separate application
Johan, we considered this approach but concluded it would require too much
re-development (more than just the database layer).
Thanks anyway.
On Tue, Feb 10, 2009 at 11:47 AM, Johan De Meersman wrote:
>
>
> On Tue, Feb 10, 2009 at 12:39 PM, Michael Addyman <
> michael.addy...
rt * Consulting * Administration
> http://www.olindata.com
>
>
>
> On Tue, Feb 10, 2009 at 12:39 PM, Michael Addyman
> wrote:
> > Thanks for your comments Mike.
> >
> > 1. The largest table has 48 columns, the second largest 20 columns, and
> the
> >
, resulting in
~3000 tables per database, and ~5 database clusters.
I think my final suggestion is the most suitable.
What would your recommendations be?
Many thanks
Michael.
On Tue, Feb 10, 2009 at 6:01 AM, mos wrote:
> At 05:03 PM 2/9/2009, Michael Addyman wrote:
>
>> Dear Geniuses,
its ratio of reading :
writing : updating.
However, this approach would require a LOT of work to re-write the
application's database layer.
What approach would be best?
Thanks again,
Michael
On Tue, Feb 10, 2009 at 6:01 AM, mos wrote:
> At 05:03 PM 2/9/2009, Michael Addyman wrote:
>
sed for its ratio of
reading : writing : updating.
However, this approach would require a LOT of work to re-write the
application's database layer.
What approach would be best?
Thanks again,
Michael
On Tue, Feb 10, 2009 at 6:01 AM, mos wrote:
> At 05:03 PM 2/9/2009, Michael Addym
Dear Geniuses,
I have an application requiring ~30 InnoDB tables, which needs to scale up
to at least 500 application instances (500 instances * ~30 tables = 15,000
tables).
Discussions in the archives suggest I would be better off having independent
databases for each of the application instance
10 matches
Mail list logo
