Colin Guthrie wrote:
> 
> Chris Mulligan wrote:
>> I agree with Noah. What's the problem this is really solving in the
>> real world? Why did you ever need a separate read database?
> 
> Not related to this general principle but i recently had to scale up a
> project I was running (totally unrelated to Trac) and if it had been
> coded with a separate read database it would have made my life a lot
> easier.... Using MySQL slaves for read operations and only hitting the
> master for writes would have made scaling up a lot easier.
> 
> So this patch does definitely solve a real life problem with regards to
> load. Sure, most trac project wont get that busy (and with trac the
> bottleneck is usually at the python end rather than the SQL end), but
> the principle is not a bad one.

Yeah, but I think this is a fairly uncommon use case. It only covers the 
case where a specific trac env is too busy for a single database server 
to keep up. And even with this patch I'm a bit skeptical that Trac will 
work very well with the asynchronous type of replication offered by 
MySQL.

So I'm "-1" since it's not work totally breaking all existing code 
(including most plugins) because of this.

/ Jonas

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Trac 
Development" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/trac-dev?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to