I guess that if I remove the EBS dependency it might work. I wonder if there
is some way keeping this.

Gluster is itself clustered or replicated so you can keep the master
redundant enough I believe.

On Tue, Mar 17, 2009 at 10:40 PM, Marten Nelson <[email protected]>wrote:

>
> Maybe this wouldn't work for your app, but I think this is where a
> cluster solution with EBS (gluster?) could help out. You store all
> application files on the cluster server. Your app servers are nodes
> and load files from the EBS attached to your cluster server. But then,
> how do you create redundancy at the cluster server? That I don't know...
>
> M
>
> On Mar 17, 2009, at 1:27 PM, Arie Fishler wrote:
>
> > I have a problem setting a procedure to synchronize my app servers
> > and avoiding downtime at the current way scalr operates.
> >
> > I am trying to figure out what will be the best way to approach this.
> >
> > The simplest example is having one app server. Let's say that my app
> > id dependent on the fact that an EBS volume is attached and mounted
> > to the app server.
> >
> > If I synchronize all, a new app server is started but not a new EBS
> > volume. Then the current running app server is taken down (=
> > downtime since the app on the new server has not yet started), the
> > EBS volume is freed and attached to the new server.
> > Only then can the application start.
> >
> > The downtime can occur exactly the same when more app servers are
> > involved. The fact that same EBS volumes are attached to the newly
> > restarted instances is a problem since my application cannot start
> > without an EBS.
> >
> > Any idea around this?
> >
> > Arie
> >
> > >
>
>
> >
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"scalr-discuss" 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/scalr-discuss?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to