Thanks for the reply . So I can just start the ovirt engine again resume the operations through ovirt engine , right ? On Fri, Jan 10, 2014 at 11:17 AM, Yair Zaslavsky <yzaslavs@redhat.com>wrote:

----- Original Message -----

...

From: "Hans Emmanuel" <hansemmanuel@gmail.com> To: users@ovirt.org Sent: Friday, January 10, 2014 7:40:23 AM Subject: [Users] Ovirt Engine single point of failure

Hi all ,

I am planning to setup an ovirt cluster with two hosts + 1 ovirt engine . But this setup have a draw back of single point of failure chance for ovirt engine. So what happens if ovirt engine goes down ? All VMs will be down ? Or it wont affect the ovirt nodes and VMs ? Please advice /

Hi, If ovirt engine crashes, your VMs will not go down.

...

-- *Hans Emmanuel*

*NOthing to FEAR but something to FEEL......*

_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users

-- *Hans Emmanuel* *NOthing to FEAR but something to FEEL......*

OK, so just so I understand this, in the described scenario of three servers: one management server/engine and two nodes, let's say the management server suffers catastrophic hard disk failure where no data can be recovered from it, nor were any backups made. Is it possible to perform a new installation of ovirt-engine, add the two existing nodes, and everything "just works"? Or would you at least need to do some reconfiguring (e.g., re-add the logical networks etc.) Basically, even though the nodes were part of the now-dead ovirt-engine, there would be no problem in getting them added in to the newly-installed ovirt-engine? -Alan

Consider trying the new hosted engine feature (http://www.ovirt.org/Features/Self_Hosted_Engine). It runs the ovirt engine as a VM inside the engine, and also provides HA solution for it. Oved ----- Original Message -----

From: "Yair Zaslavsky" <yzaslavs@redhat.com> To: "Alan Murrell" <lists@murrell.ca> Cc: users@ovirt.org Sent: Friday, January 10, 2014 9:17:53 AM Subject: Re: [Users] Ovirt Engine single point of failure

Alan, IMHO this is not the scenario described in the original question - or maybe I did not understand well the original question? I assume the original question is about a scenario where engine restarts, and not about a catastrophic failure as you describe here.

----- Original Message -----

...

From: "Alan Murrell" <lists@murrell.ca> To: users@ovirt.org Sent: Friday, January 10, 2014 9:01:14 AM Subject: Re: [Users] Ovirt Engine single point of failure

OK, so just so I understand this, in the described scenario of three servers: one management server/engine and two nodes, let's say the management server suffers catastrophic hard disk failure where no data can be recovered from it, nor were any backups made.

Is it possible to perform a new installation of ovirt-engine, add the two existing nodes, and everything "just works"? Or would you at least need to do some reconfiguring (e.g., re-add the logical networks etc.)

Basically, even though the nodes were part of the now-dead ovirt-engine, there would be no problem in getting them added in to the newly-installed ovirt-engine?

-Alan

_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users

_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users

For now only for testing purposes, and considering it to be used in the near future once it is more tested/stable. It will be released only on oVirt 3.4 alpha, and I guess it should be complete and tested in 3.4. cc-ing Doron, which is pushing this feature. He should know best about its current status. Thank you, Oved ----- Original Message -----

From: "Sven Kieske" <S.Kieske@mittwald.de> To: users@ovirt.org Sent: Monday, January 13, 2014 10:46:56 AM Subject: Re: [Users] Ovirt Engine single point of failure

Hi,

do you really advise to use a feature which is not 100% implemented and is just available in an alpha release for a production setup?

Or is my assumption incorrect and this feature is already 100% supported? This would be great news.

Am 12.01.2014 07:15, schrieb Oved Ourfalli:

...

Consider trying the new hosted engine feature (http://www.ovirt.org/Features/Self_Hosted_Engine). It runs the ovirt engine as a VM inside the engine, and also provides HA solution for it.

Oved

-- Mit freundlichen Grüßen / Regards

Sven Kieske

Systemadministrator Mittwald CM Service GmbH & Co. KG Königsberger Straße 6 32339 Espelkamp T: +49-5772-293-100 F: +49-5772-293-333 https://www.mittwald.de Geschäftsführer: Robert Meyer St.Nr.: 331/5721/1033, USt-IdNr.: DE814773217, HRA 6640, AG Bad Oeynhausen Komplementärin: Robert Meyer Verwaltungs GmbH, HRB 13260, AG Bad Oeynhausen _______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users

On 01/10/2014 09:01 AM, Alan Murrell wrote:

OK, so just so I understand this, in the described scenario of three servers: one management server/engine and two nodes, let's say the management server suffers catastrophic hard disk failure where no data can be recovered from it, nor were any backups made.

Is it possible to perform a new installation of ovirt-engine, add the two existing nodes, and everything "just works"? Or would you at least need to do some reconfiguring (e.g., re-add the logical networks etc.)

Basically, even though the nodes were part of the now-dead ovirt-engine, there would be no problem in getting them added in to the newly-installed ovirt-engine?

-Alan

_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users

for now, you should backup the engine on a regular basis. once we close the gap of 'import existing data domain' will be easier to recover even without an engine backup (you'd still need to add the hosts of course).