Thank you, that's very helpful. Am I right to infer from the GUI and the existing documentation that synchronization of the cluster members' PacketFence configuration (e.g.: /usr/local/pf/conf/*) must be achieved outside of the configuration of PacketFence (through rsync, or what have you)? I see mention of this in several packetfence-users postings--and several approaches are discussed--but I wanted to be sure not to invent something myself, if configuration sync had already been implemented in the code, or if a "best practice" approach were already documented.
On Mon, Oct 26, 2015 at 6:21 PM, Louis Munro <[email protected]> wrote: > The guide is intended to cover Active/Active setups. > > You are correct that regardless of the approach you take, a database has > to be shared somehow. > > In our experience DRBD works very well and is probably the least > troublesome part of a PacketFence cluster. > But any approach which allows you to lose one node out of a cluster is > probably good enough redundancy for most people. > > You could have your database in a master/slave configuration. > That would work just as well although it may require more scripting or > manual intervention. > > There is no one-size fits all redundant deployment. > I encourage you to try another approach if you think it will suit you > better and share your experience with the list. > > Regards, > -- > Louis Munro > [email protected] :: www.inverse.ca > +1.514.447.4918 x125 :: +1 (866) 353-6153 x125 > Inverse inc. :: Leaders behind SOGo (www.sogo.nu) and PacketFence ( > www.packetfence.org) > > On Oct 26, 2015, at 18:17 , Dale Whiteaker-Lewis <[email protected]> > wrote: > > I'm confused as to whether > http://www.packetfence.org/downloads/PacketFence/doc/PacketFence_Clustering_Guide-5.4.0.pdf > represents the instructions for Active/Passive setup, or only Active/Active > setup. I am interested in taking the least complicated approach--in > addition to the standalone configuration described in the other guides--to > having a fail-over option that can be switched over on the order of a > couple of minutes, not instantaneously, nor taking hours. > > I think that guide is telling me that even in an active/passive mode, the > active and passive servers need to share a database, and that to do so > reliably requires something like DRBD. Is this a correct interpretation? > Because, I'd really like to avoid kernel mods and such, if active/passive > can be done without them. > > I've searched the mailing list, and I see relatively few references to > DRBD, but several references, including in May of this year, to > synchronizing state in other, scripted ways. > > ------------------------------------------------------------------------------ > _______________________________________________ > PacketFence-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/packetfence-users > > > > > ------------------------------------------------------------------------------ > > _______________________________________________ > PacketFence-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/packetfence-users > >
------------------------------------------------------------------------------
_______________________________________________ PacketFence-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/packetfence-users
