Global Presence use cases - Fluid Topics - 4.0

Fluid Topics Installation Guide

Operating system
RHEL
Category
Reference Guides
Audience
public
Version
4.0

The following scenarios demonstrate how Global Presence improves resiliency while protecting data and ensuring content availability.

Scenario 1

Event

One of the MongoDB servers of an FT cluster is unavailable.

Impact

No impact.

Outcome

The node synchronizes as soon as the MongoDB server is available again.

Scenario 2

Event

Two of the MongoDB servers of an FT cluster in the same region are unavailable.

Impact

No impact.

Outcome

The primary server in Europe performs all read and write operations until one of the edge servers is available again.

Scenario 3

Event

One of the MongoDB servers of a Nerada cluster is unavailable.

Impact

No impact.

Outcome

The node synchronizes as soon as the lost MongoDB server is available again.

Scenario 4

Event

Two of the MongoDB servers of a Nerada cluster in the same region are unavailable.

Impact

Since there is no longer a MongoDB server available locally, the closest remote server is used.

Outcome

Portal performance is degraded locally until at least one of the servers is available again.

Scenario 5

Event

An edge data center fails to establish a VPN connection to the primary data center in Europe.

Impact

There may be a short interruption in service for the edge data center that loses its VPN connection, and users in the region may temporarily be unable to upload content. However, if a user does manage to upload content, indexation remains available. Administration interfaces that require a VPN connection (BO, indexing servers, Analytics) are not available. There is no effect for the other data centers.

Outcome

In the region that experienced an issue, service returns to normal as the VPN connection is reestablished.

Scenario 6

Event

The VPN connection is lost in the primary data center and all edge data centers simultaneously.

Impact

Portal performance is severely degraded. It is still possible to browse content, but it is impossible to upload, and indexation is not available.

Outcome

Portal performance returns to normal as soon as the primary data center reestablishes its VPN connection. Components are restored in the AWS cloud infrastructure.