Re: 2 Real Servers & 2 Directors with only 2 machines

On Tue, 2004-02-03 at 15:03, Mike McLean wrote:
> Edward Croft wrote:
> > Greg, with Sebastien's permission, I submitted his changes to Red Hat,
> > but I think that they have all but abandoned piranha in favor of Cluster
> > Manager. Shame too, since piranha seems to work fairly well for
> > failover. 
> 
> Not abandoned.  Piranha has two modes, FOS and LVS.  When Cluster 
> Manager was added to the distro, we had two different products for 
> failover.  Since Cluster Manager is superior in the failover department 
> it became the preferred solution and Piranha's FOS mode became deprecated.
> 
> However, Piranha's LVS mode (which really is its main feature) is still 
> the preferred way to perform load balancing.  Failover between the two 
> directors is also still supported.
> 
> I cannot find an RFE in bugzilla regarding the failover script patch. 
> If this applies to director failover I will consider putting it in.
> 

The code that Sebastien modifed was to allow FOS to use scripts. We use
this extensively. Piranha is simple and it works. Especially with the
patch. My understanding was that Cluster Manager did LVS. With our
situation, the patch worked great. It allows more indepth checking for
failover. 

-- 
Edward M. Croft
Sr. Systems Engineer
Open Ratings, Inc.
200 West Street
Waltham, MA 02451-1121