cloudstack-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From benoit lair <kurushi4...@gmail.com>
Subject Re: [ACS 43] [MGMT Server][High Availability]
Date Mon, 13 Oct 2014 09:50:19 GMT
Ok, you are using the Xen hypervisor, not Xenserver.

I would like to achieve my replication with drbd and xenserver.

But i was not thinking about doing it at the hypervisor level but in the vm
level.



2014-10-10 15:40 GMT+02:00 France <mailinglists@isg.si>:

> For ACS we are using XS 6.0.2+hotfixes.
> For the “old cloud” where management servers reside in a form of XEN PV
> machines we are using RHEL.
> OCFS2 is not needed when virtual machine uses DRBD device directly. We
> have one DRBD device per each VM.
>
> Regards,
> F.
>
> On 09 Oct 2014, at 16:13, benoit lair <kurushi4000@gmail.com> wrote:
>
> > Hello,
> >
> > Ok, bu i would like to have a separated mysql db server and mgmt servers.
> > Are you using KVM or xenserver ?
> > With your active/active drbd cluster, are you using ocfs2 ?
> >
> > 2014-10-08 12:34 GMT+02:00 France <mailinglists@isg.si>:
> >
> >> Put the ACS java app in the same server and it will always get working
> >> mysql server when it is working.
> >> Also I suggest you start using pacemaker, corosync or cman.
> >>
> >> My management server is actually a virtual instance on RHEL
> active/active
> >> drbd cluster (so live migration works).
> >>
> >> If you want to test how it behaves, stop mysql and check for yourself. I
> >> highly doubt anyone has tested it yet.
> >>
> >> Regards,
> >> F.
> >>
> >> On 08 Oct 2014, at 11:37, benoit lair <kurushi4000@gmail.com> wrote:
> >>
> >>> Hello Folks,
> >>>
> >>>
> >>> I'm trying new HA implementation of my mgmt server.
> >>>
> >>> I'm looking for HA for the mysql server.
> >>>
> >>> Could it be problematic if i install a drbd active/passive mysql
> cluster
> >> ?
> >>> (drbd & heartbeat)
> >>>
> >>> The reason is because if i a have a fail of my primary server (and so
> the
> >>> cluster doing a failover and transmiting its VIP (heartbeat) to the
> other
> >>> node), the mysql server doesn't respond during few seconds (due to the
> >>> deadtime parameter).
> >>>
> >>> So is this scenario problematic for the integrity of the management
> >> server ?
> >>>
> >>>
> >>> Thanks for your responses.
> >>
> >>
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message