[ovirt-users] Engine HA?

Thu Jun 12 10:34:03 UTC 2014

I have already done a POC.
I'm going eventually to post inside the mailing list the gory details so
that you could check and give feedback and then I go contribute in the wiki.
Let me recollect the doc with the steps.
Basic system was CentOS 6.5 with Pacemaker and cman.
The filesystem resources (for postgres, etc..) were managed in drbd
primary/secondary.

At the end you would have somethng like this:

[root at ovirteng01 ~]# crm_mon -1
Last updated: Thu Jun 12 12:21:13 2014
Last change: Fri Mar 14 15:30:57 2014 via cibadmin on
ovirteng01.localdomain.local
Stack: cman
Current DC: ovirteng02.localdomain.local - partition with quorum
Version: 1.1.10-14.el6_5.2-368c726
2 Nodes configured
13 Resources configured

Online: [ ovirteng01.localdomain.local ovirteng02.localdomain.local ]

 Master/Slave Set: ms_OvirtData [OvirtData]
     Masters: [ ovirteng02.localdomain.local ]
     Slaves: [ ovirteng01.localdomain.local ]
 Resource Group: ovirt
     ip_OvirtData       (ocf::heartbeat:IPaddr2):       Started
ovirteng02.localdomain.local
     lvm_ovirt  (ocf::heartbeat:LVM):   Started ovirteng02.localdomain.local
     fs_OvirtData       (ocf::heartbeat:Filesystem):    Started
ovirteng02.localdomain.local
     pgsql_OvirtData    (lsb:postgresql):       Started
ovirteng02.localdomain.local
     p_lsb_nfs  (lsb:nfs):      Started ovirteng02.localdomain.local
     p_exportfs_root    (ocf::heartbeat:exportfs):      Started
ovirteng02.localdomain.local
     p_exportfs_iso     (ocf::heartbeat:exportfs):      Started
ovirteng02.localdomain.local
     ovirt-engine       (lsb:ovirt-engine):     Started
ovirteng02.localdomain.local
     ovirt-websocket-proxy      (lsb:ovirt-websocket-proxy):    Started
ovirteng02.localdomain.local
     httpd      (ocf::heartbeat:apache):        Started
ovirteng02.localdomain.local

or with pcs commands:

[root at ovirteng01 ~]# pcs resource show --groups
ovirt: ip_OvirtData lvm_ovirt fs_OvirtData pgsql_OvirtData p_lsb_nfs
p_exportfs_root p_exportfs_iso ovirt-engine ovirt-websocket-proxy httpd

[root at ovirteng01 ~]# pcs resource show
 Master/Slave Set: ms_OvirtData [OvirtData]
     Masters: [ ovirteng02.localdomain.local ]
     Slaves: [ ovirteng01.localdomain.local ]
 Resource Group: ovirt
     ip_OvirtData       (ocf::heartbeat:IPaddr2):       Started
     lvm_ovirt  (ocf::heartbeat:LVM):   Started
     fs_OvirtData       (ocf::heartbeat:Filesystem):    Started
     pgsql_OvirtData    (lsb:postgresql):       Started
     p_lsb_nfs  (lsb:nfs):      Started
     p_exportfs_root    (ocf::heartbeat:exportfs):      Started
     p_exportfs_iso     (ocf::heartbeat:exportfs):      Started
     ovirt-engine       (lsb:ovirt-engine):     Started
     ovirt-websocket-proxy      (lsb:ovirt-websocket-proxy):    Started
     httpd      (ocf::heartbeat:apache):        Started

the share file system is something like
/dev/mapper/VG_OVIRT-LV_OVIRT  5.0G  203M  4.5G   5% /shared

and has structure like
[root at ovirteng02 ~]# ll /shared
total 28
drwxr-xr-x. 6 root root  4096 Mar  1 14:11 etc
drwx------. 2 root root 16384 Mar  1 10:29 lost+found
drwxr-xr-x. 5 root root  4096 Mar  5 17:10 nfs_info
drwxr-xr-x. 3 root root  4096 Mar  1 13:52 var

[root at ovirteng02 ~]# ll /shared/etc/
total 16
drwxr-xr-x.  4 root root 4096 Mar  1 17:30 httpd
drwxr-xr-x. 15 root root 4096 Mar  1 13:58 ovirt-engine
drwxr-xr-x.  2 root root 4096 Mar  1 13:59 ovirt-engine-setup.conf.d
drwxr-xr-x.  3 root root 4096 Mar  1 14:03 pki

[root at ovirteng02 ~]# ll /shared/var/
total 4
drwxr-xr-x. 6 root root 4096 Mar  6 01:52 lib

[root at ovirteng02 ~]# ll /shared/var/lib/
total 16
drwxr-xr-x. 3 root     root     4096 Mar  5 18:58 exports
drwxr-xr-x. 4 root     root     4096 Mar 14 15:28 nfs
drwxr-xr-x. 8 ovirt    ovirt    4096 Mar 12 00:30 ovirt-engine
drwx------. 4 postgres postgres 4096 Mar  1 12:49 pgsql

more details in following e-mail.
Let me know if it can be of interest.

I tried some things like forced panic, relocation while webadmin open,
etc... and it seemed consistent.

Gianluca

On Thu, Jun 12, 2014 at 11:52 AM, Itamar Heim <iheim at redhat.com> wrote:

> On 06/12/2014 05:44 AM, KevinTang at umac.mo wrote:
>
>> Dear all,
>>
>> If not using self hosted-engine solution, my Engine will install in
>> dedicate machine. How should I make the engine become HA?
>>
>
> easiest would be to run it as a VM on a clustering solution like Red Hat
> Cluster Suite or similar.
>
>
>
>> Thanks
>>
>> Best Regards,
>> Kevin Tang
>> ------------------------------------------------------------
>> --------------------------------------------------------
>> AMSV - STATE KEY LABORATORY OF ANALOG AND MIXED-SIGNAL VLSI
>> University of Macau
>> Tel: (+853) 8397-8035
>>
>>
>>
>> From: Itamar Heim <iheim at redhat.com>
>> To: KevinTang at umac.mo, users at ovirt.org
>> Date: 06/12/2014 04:07 AM
>> Subject: Re: [ovirt-users] Engine HA?
>> ------------------------------------------------------------------------
>>
>>
>>
>>
>> On 06/11/2014 04:16 AM, KevinTang at umac.mo wrote:
>>  > Dear all,
>>  >
>>  > I know oVirt Host (VM) can be HA and cluster, all VM can be migrate
>>  > between all operational Node.
>>  >
>>  > Since the main oVirt Engine is standalone, event shelf-hosted engine,
>> it
>>  > still is a single Engine.
>>  >
>>  > I want to ensure my Engine can be HA, how should I do? Do I need to
>>  > create a Cluster Linux first?
>>  >
>>
>> hosted-engine has built-in HA mechanism for the engine, so if the host
>> its running on has an issue, another host will launch the hosted engine.
>>
>>
>>
>>
> _______________________________________________
> Users mailing list
> Users at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/users/attachments/20140612/6776d13c/attachment-0001.html>