That makes sense.  I have some experience with clustered filesystems
(GFS, GFS2) and DRBD, so I don't think that will be too hard to manage.

I'm assuming that Trac can deal with multiple concurrent writers trying
to modify these resources.  I don't have any good synthetic benchmarks
that I could run against Trac to look for concurrency problems on my
own, and I'd rather not wait til the system is under load to find out
about them, you know?

-Ryan


Evan Dower wrote:
> Your trac environment directory would also need to be shared (via NFS
> perhaps), or read-only.  Otherwise your environment config might get
> out of sync on different boxes, which seems like it would cause
> problems.
> 
> On Sun, Feb 15, 2009 at 3:44 AM, Noah Kantrowitz <[email protected]> wrote:
>> Yes, this is fine. When using mod_py with mpm_prefork, you end up with
>> lots of worker processes accessing the same environment anyway :-)
>>
>> --Noah
>>
>> On Feb 14, 2009, at 11:54 AM, Ryan B. Lynch wrote:
>>
>>> I posted this question to the users list, a while ago, and didn't
>>> see a
>>> response, so I figured the development list might have an answer.
>>>
>>> Is it possible to point multiple web server front-ends at the same
>>> Trac
>>> database (using MySQL) and have consistent operations?  I'm using
>>> Apache
>>> with mod_python.  All of the front-end machines would be in
>>> simultaneous
>>> use.
>>>
>>> -Ryan
>>>
>>
> 
> > 

--~--~---------~--~----~------------~-------~--~----~
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