[rdo-list] tripleo ha topology

[Bruce Tan]

>From a Hi-Availability point of view, I would say that
1. HAProxy is the mechanism to manage multiple copies of actual openstack services (e.g. MariaDB).
So when one service stops, HAProxy will load-balance or fail-over to the rest instance(s), and such services will not be single point of failure.

2. Similarly, we need a fail-over mechanism for HAProxy, or the HAProxy will be a single point of failure. And hence the pacemaker.
We use pacemaker to ensure that when one HAProxy service (instance) fails, some other instance will take over.

Best regards,
Bruce Tan
> On 15 Dec 2016, at 16:54, Raoul Scarazzini <rasca at redhat.com> wrote:
> 
> On 15/12/2016 09:42, Dotan, Alon wrote:
>> So if understood correctly HAproxy is basically a component in pcs?
> 
> It is a systemd resource managed by Pacemaker.
> 
> -- 
> Raoul Scarazzini
> rasca at redhat.com
> 
>> -----Original Message-----
>> From: Raoul Scarazzini [mailto:rasca at redhat.com] 
>> Sent: Thursday, December 15, 2016 10:33 AM
>> To: Dotan, Alon <alon.dotan at hpe.com>; rdo-list at redhat.com
>> Subject: Re: [rdo-list] tripleo ha topology
>> 
>> On 14/12/2016 22:59, Dotan, Alon wrote:
>>> Hey,
>>> Some can provide link or explanation about the use of haproxy and 
>>> pacemaker together in tripleo?
>>> Why need pcs if there is galera and haproxy?
>> 
>> Galera is managed via a specific resource agent which takes care of all the stuff needed to start, stop and, most of all, recover. It automates a lot of tasks which otherwise will be manual.
>> 
>> Haproxy is managed by Pacemaker because of the Vips. There are specific constraints that ensures that whether a Vip change the host, haproxy service will follow.
>> 
>> Hope this helps,
>> 
>> --
>> Raoul Scarazzini
>> rasca at redhat.com
>> 
> 
> 
> _______________________________________________
> rdo-list mailing list
> rdo-list at redhat.com
> https://www.redhat.com/mailman/listinfo/rdo-list
> 
> To unsubscribe: rdo-list-unsubscribe at redhat.com
>