Hi Oliver,
The soft HA (e.g.: you connect to the stand-by ceph-mgr) will still work
even with IP addresses, as it uses HTTP 3xx redirection.
For having hard HA, there are many ways: L3 (keepalived), L4-7 (HAproxy,
DNS round-robin, etcd/consul/etc.). As cephadm allows you to easily deploy
"keepalived + haproxy" ("ingress" service) in front of the Dashboard, you
could give that a try (I haven't tested that yet, but it's on our roadmap
to explore x3 HA for the Dashboard + monitoring stack).
Kind Regards,
Ernesto
On Wed, Sep 1, 2021 at 10:47 AM Oliver Weinmann <oliver.weinmann@xxxxxx>
wrote:
> Hi Ernesto,
>
> thanks for pointing me in the right direction. I have also found a bug
> report for the issue:
>
> https://tracker.ceph.com/issues/52002
>
> But maybe I'm misunderstanding this whole network setup. Let's say I'm
> starting to get confused.
>
> For example:
>
> I have started a very small deployment of a 3 node (4 nodes if you include
> the cephadm bootstrap node) cluster. I wanted to do this by best practice,
> and so I created a bond (XOR) for my 10G nics and a VLAN for public (
> 192.168.30.0/24) and one for cluster (192.168.41.0/24). I would like to
> add additional networks (VLANs) later on for varios service e.g. ISCSI etc.
> The hosts FQDN are not pointing to 192.168.30.x but to another network
> where all our hosts reside that are deployed via kickstart. Do I have to
> ensure that the ceph hosts FQDN points to the 192.168.30 network or is this
> just for conveniance? I don't mind using the IP of the active mgr to visit
> the dashboard, but I guess this kind of breaks the idea of having
> redundancy in case the active mgr fails. I would then have to check who is
> the active mgr and use it's 192.168.30.x ip adress.
>
> Best Regards,
> Oliver
>
> Am 31. August 2021 um 12:16 schrieb Ernesto Puerta <epuertat@xxxxxxxxxx>:
>
> Hi Oliver,
>
> This issue has already been discussed in this mailing list ([1]
> <https://lists.ceph.io/hyperkitty/list/ceph-users@xxxxxxx/thread/LWQKSRTO5HOAABFZDA26WGF3TL2JHLVI/#LWQKSRTO5HOAABFZDA26WGF3TL2JHLVI>
> and [2]
> <https://lists.ceph.io/hyperkitty/list/ceph-users@xxxxxxx/thread/T2WD24E4BLHG2DYH4BEGSD7XT2XTHFOP/#T2WD24E4BLHG2DYH4BEGSD7XT2XTHFOP>).
> There you can find the details and steps to deal with it.
>
> Kind Regards,
> Ernesto
>
>
> On Mon, Aug 16, 2021 at 12:51 PM Oliver Weinmann <oliver.weinmann@xxxxxx>
> wrote:
>
>> Dear All,
>>
>>
>>
>> after a very smooth upgrade from 16.2.4 to 16.2.5 (CentOS 8 Stream), we
>> are no longer able to access the dashboard. The dashboard was accessible
>> before the upgrade. I googled and found a command to change the listening
>> IP for the dashboard, but I wonder why the upgrade should have changed this?
>>
>>
>>
>> https://people.redhat.com/bhubbard/nature/default/mgr/dashboard/
>>
>>
>> ceph mgr services
>> {
>> "dashboard": "https://192.168.30.198:8443/";,
>> "prometheus": "http://192.168.30.198:9283/";
>> }
>>
>>
>>
>> Just for testing I change the IP of the dashboard and now I can access it
>> and log in but then there is no data, no menu display. I assume this is
>> because prometheus IP also needs to be changed? But before I do this, I
>> would like to understand why this was changed during the upgrade.
>>
>>
>>
>> Best Regards,
>>
>> Oliver
>>
>>
>> _______________________________________________
>> ceph-users mailing list -- ceph-users@xxxxxxx
>> To unsubscribe send an email to ceph-users-leave@xxxxxxx
>>
>
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
