I'm alright with that for the time being :-)

I was also interested in how I would get my two PF instances to share the
same data, so if there's a failover the data would transfer over? I may be
going in circles here, and the documentation originally provided contains
the answers. I looked through it best I could, but have a hard time fully
understanding it..

On Wed, Apr 24, 2019 at 10:09 AM Ludovic Zammit <[email protected]> wrote:

> Hello,
>
> Most of the time switches support that feature where when the first server
> is not responsive it will forward the radius request to the second one.
>
> It will fail after fews radius failure. Most of the time it can be tuned
> on the switch radius settings.
>
> It’s not a true HA in a way that you would few minutes of down time if one
> side fail.
>
> Thanks,
>
>
> Ludovic [email protected] ::  +1.514.447.4918 (x145) ::  www.inverse.ca
> Inverse inc. :: Leaders behind SOGo (http://www.sogo.nu) and PacketFence 
> (http://packetfence.org)
>
>
>
>
>
> On Apr 24, 2019, at 9:04 AM, Stuart Gendron <[email protected]>
> wrote:
>
> Thank you.
>
> Is there a way to setup HA on the switch side (have two RADIUS servers,
> try the next one if one fails) and then have a remote database that two
> PacketFence instances can look at? Basically have two PF servers that are
> in sync?
>
> I'm looking through the guide and it's going over my head a bit..
>
> On Tue, Apr 23, 2019 at 4:02 PM Ludovic Zammit <[email protected]> wrote:
>
>> Hello Stuart,
>>
>> The answer is pretty simple.
>>
>> You are looking at two different version of PacketFence that uses two
>> different datable technology.
>>
>> PF 6.5.1 relied on mysql version with corosync + pacemaker + drbd and PF
>> 8.X relies on MariaDB Galera cluster for HA.
>>
>> With PF 8 version you need 2 server + 1 for the qorum.
>>
>> Thanks,
>>
>>
>> Ludovic [email protected] ::  +1.514.447.4918 (x145) ::  
>> www.inverse.ca
>> Inverse inc. :: Leaders behind SOGo (http://www.sogo.nu) and PacketFence 
>> (http://packetfence.org)
>>
>>
>>
>>
>>
>> On Apr 23, 2019, at 3:50 PM, Stuart Gendron via PacketFence-users <
>> [email protected]> wrote:
>>
>> Hey there,
>>
>> Sorry for the easy question, however I'm getting confused when looking
>> around on how to configure PacketFence for High Availability.
>>
>> The more recent instructions:
>>
>> https://packetfence.org/doc/PacketFence_Clustering_Guide.html
>>
>> state that you need a minimum of 3 PF instances.
>>
>> Whereas the older instructions
>>
>>
>> https://packages.inverse.ca/PacketFence/doc/PacketFence_Clustering_Guide-6.5.1.pdf
>>
>> state you only need 2 PF instances.
>>
>> 2 instances would be ideal, as 3 is definitely overkill for our
>> environment.
>>
>> Was wondering if I'd run into serious issues with following instructions
>> for 6.5.1, or if there's more up to date instructions that only require 2
>> instances.
>>
>> Thanks again for everyone's help. Have saved me a great deal of trouble.
>>
>> --
>>
>> *Stuart Gendron*
>> IT Support Specialist
>>
>> *You.i Labs*
>> 307 Legget Drive, Kanata, ON, K2K 3C8
>> <https://maps.google.com/?q=307+Legget+Drive,+Kanata,+ON,%C2%A0K2K+3C8&entry=gmail&source=g>
>> t (613) 228-9107 x258 | c (613) 697-6853
>> _______________________________________________
>> PacketFence-users mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/packetfence-users
>>
>>
>>
>
> --
>
> *Stuart Gendron*
> IT Support Specialist
>
> *You.i Labs*
> 307 Legget Drive, Kanata, ON, K2K 3C8
> <https://maps.google.com/?q=307+Legget+Drive,+Kanata,+ON,%C2%A0K2K+3C8&entry=gmail&source=g>
> t (613) 228-9107 x258 | c (613) 697-6853
>
>
>

-- 

*Stuart Gendron*
IT Support Specialist

*You.i Labs*
307 Legget Drive, Kanata, ON, K2K 3C8
<https://maps.google.com/?q=307+Legget+Drive,+Kanata,+ON,%C2%A0K2K+3C8&entry=gmail&source=g>
t (613) 228-9107 x258 | c (613) 697-6853
_______________________________________________
PacketFence-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/packetfence-users

Reply via email to