High availability or resiliant hosting of Geonode

[Marcathian Alexander]

Hi Every one,
We have installed and launched our geonode platform here in Saint Lucia (http://sling.gosl.gov.lc).
I would like to get your thoughts, comments and suggestions on the issue presented below.

At the moment the system is physically hosted on the island.  We would like to have a more robust hosting arrangement. 
This arrangement would allow for access to the system by off-island users (e.g international aid agencies)
 in the event of a catastrophe that hinders communications with the island.
It should also allow for access by on-island users after communications with the outside world has been interrupted.

So I am thinking that at least two physical points are needed (1) on the island of Saint Lucia and (2) some where away from the island.
Both systems would need to allow users to have read/write access to the data as their permissions allow, they also need to be kept synchronized.

Please let me hear your thoughts and suggestions on ways to achieve this.
If any one has had experience with a similar arrangement, I would like to hear about it.

Regards.


Marcathian Alexander
Information Technology Officer I
Ministry of Physical Development, Housing and Urban Renewal
Greaham Louisy Administrative Building,
P.O. Box 709, The Waterfront, Castries,

Tel. (758) 468-4434;
Mobile: (758) 719-2248;
Fax (758) 452-2506;
email: macale...@gosl.gov.lc
       malex...@sde.gov.lc

[Gerry George]

To be able to determine the redundancy options, please provide details
of system configuration: Server OS, application backends (if any),
access type (direct, web, TCP port or whatever).

Without any further information to make a proper determination, I can
suggest the option of a mirrored (or synchronized) service between
the two (or more) servers to provide for redundancy and authenticated
user access. However, the configuration and the possibility of doing
this depends on the system and program configuration.

Gerry George
DigiSolv, Inc.

[jackrs...@gmail.com]

Hi, I know I'm resurrecting an old thread, apologies, but it's relevant.

I'm working with a national government with an extremely similar use case to the OP. Presently, they have a GeoNode instance installed locally and accessible publicly from the internet. This is fragile due to infrastructure weaknesses on premises (e.g. unreliable power). I'm considering recommending they move to an instance of GeoNode installed in the cloud (perhaps AWS, but a managed solution would be ideal). This instance would be locally mirrored on site using their current hardware. The local mirror is essential redundancy to accommodate scenarios where the internet may go down (e.g. emergency management). The local mirror should remain relatively in-sync with the cloud master, but infrequent synching, perhaps once a day, should sufficiently cover the anticipated usage patterns.

I figure the devil is in the details here, as in a genuine emergency, if the internet is down and changes are being made locally directly on the mirror, once internet is back, the synching strategy with the master suddenly gets complicated.

I'm not able to find examples of this type of setup in the docs or otherwise. As I'm not a sys admin, I may be using the wrong search terms. Can someone point me toward resources that provide further details on implementing this type of strategy and best practices? If anyone has additional relevant insights they can share, I'd greatly appreciate it.

Appreciate your help.