Failover for Redis Enterprise with Active-geo-Replication

Question

Failover for Redis Enterprise with Active-geo-Replication

Giden, Burcu 20

using terraform I have created two redis enterprise clusters in paired regions ( West and northEurope).

xx-redis-primary in WestEurope

xx-redis-secondary in NorthEurope

And linked them together using active geo-replication... /Groupnickame xxRedisGeoGroup

now I want to trigger manual failover..

Scenario: WestEurope is unavailable

step1) Force Unlink

az redisenterprise database force-unlink \
--unlink-ids "/subscriptions/<subscriptionid>/resourceGroups/<resourcegroupname>/providers/Microsoft.Cache/redisEnterprise/
--resource-group <resourcegroupname> \
--cluster-name

or

Invoke-AzRedisEnterpriseCacheForceDatabaseUnlink -ResourceGroupName <resourcegroupname> -ClusterName "xx-redis-primary" -Id @("/subscriptions/<subscriptionid>/resourceGroups/<resourcegroupname>/providers/Microsoft.Cache/redisEnterprise/xx-redis-secondary/databases/default")

with this step I am able to unlink west europe and north europe... what happens actually in that case? the only active one is the north europe? or both are still active however no replication in between? what is the behaviour in this case?

Step2) re-Link them

once the westeurope is available again how can I link them again.. using az redisterprise database create? if so it isnt working...or I dont understand the syntax

az redisenterprise database create \
--cluster-name xx-redis-primary
--resource-group <resourcegroupname>
--group-nickname xxRedisGeoGroup
--linked-databases "resource= \
subscriptions/<subscriptionid>/resourceGroups/<resourcegroupname>/providers/Microsoft.Cache/redisEnterprise/xx-redis-primary/databases/default.force-link" \
--linked-databases "resource= \
subscriptions/<subscriptionid>/resourceGroups/<resourcegroupname>/providers/Microsoft.Cache/redisEnterprise/xx-redis-secondary/databases/default.force-link" \
--client-protocol Encrypted \  
--clustering-policy EnterpriseCluster \
--eviction-policy AllKeysLRU \
--port 10000

or

Invoke-AzRedisEnterpriseCacheForceDatabaseLinkToReplicationGroup  -ClusterName "xx-redis-primary"  -ResourceGroupName "resourcegroupname"  -DatabaseName "default"  -GroupNickname "xxRedisGeoGroup"  -LinkedDatabase @(@{ResourceId="/subscriptions/subscriptionid/resourceGroups/resourcegroupname/providers/Microsoft.Cache/redisEnterprise/xx-redis-primary/databases/default"}, @{ResourceId="/subscriptions/
subscriptionid/resourceGroups/resourcegroupname

however it doesnt function...

Giden, Burcu 20 Reputation points

2025-03-01T10:35:25.7933333+00:00
Behavior After Force Unlink: After unlinking, is NorthEurope still active (as the primary), and will it serve all traffic until the failover completes?

Do I need to set up Azure Traffic Manager with a priority routing profile for proper failover handling between WestEurope and NorthEurope? Would this ensure proper routing when failover occurs, and traffic needs to be automatically directed to NorthEurope (secondary) during an outage in WestEurope?

Re-linking Issues: Can anyone guide me on how to correctly relink the Redis Enterprise clusters using Azure CLI or PowerShell? Is there a different approach to use for the re-linking process?
Sai Raghunadh M 2,880 Reputation points Microsoft External Staff

2025-03-03T01:29:36.5033333+00:00
Hi @Giden, Burcu

Thanks for the Question and using Microsoft Q&A

Behavior After Force Unlink:

After you force unlink the Redis Enterprise clusters, the NorthEurope cluster will continue to be active and serve all traffic until the failover completes. The WestEurope cluster will be unavailable, and there will be no replication between the two clusters during this period.

Setting Up Azure Traffic Manager:

Yes, setting up Azure Traffic Manager with a priority routing profile is a good approach for handling failover between WestEurope and NorthEurope. This ensures that traffic is automatically directed to the NorthEurope cluster (secondary) during an outage in WestEurope. The priority routing method allows you to define a primary endpoint and one or more backup endpoints. When the primary endpoint is unavailable, traffic is routed to the backup endpoint.

Re-linking Redis Enterprise Clusters:

Once the WestEurope is available, try following the same Azure CLI script to relink the clusters that you have mentioned above.

az redisenterprise database create \ --cluster-name xx-redis-primary \ --resource-group <resourcegroupname> \ --group-nickname xxRedisGeoGroup \ --linked-databases "resource=/subscriptions/<subscriptionid>/resourceGroups/<resourcegroupname>/providers/Microsoft.Cache/redisEnterprise/xx-redis-primary/databases/default.force-link" \ --linked-databases "resource=/subscriptions/<subscriptionid>/resourceGroups/<resourcegroupname>/providers/Microsoft.Cache/redisEnterprise/xx-redis-secondary/databases/default.force-link" \ --client-protocol Encrypted \ --clustering-policy EnterpriseCluster \ --eviction-policy AllKeysLRU \ --port 10000

Please refer below Documentations that might help you:

https://learn.microsoft.com/en-us/azure/traffic-manager/traffic-manager-configure-priority-routing-method

https://learn.microsoft.com/en-us/cli/azure/redisenterprise?view=azure-cli-latest

https://learn.microsoft.com/en-us/cli/azure/redisenterprise?view=azure-cli-latest

Hope this helps. Do let us know if you any further queries.
Giden, Burcu 20 Reputation points

2025-03-03T10:32:44.04+00:00

thanks for checking this.. however you have written exactly the same cli command that I have used to re-link them again- there is no difference in between... and it doesnt function..

the error I receive is "Failed to parse '--linked-databases' argument: "Model 'AAZObjectArg' has no field named 'resource'""
Giden, Burcu 20 Reputation points

2025-03-03T10:35:20.0133333+00:00

as written in my main question the az cli command is 1-1 the same with the one you wrote and it doesnt function !!! see the picture.. would appreciate any suggestions..
Sai Raghunadh M 2,880 Reputation points Microsoft External Staff

2025-03-03T14:01:15.8266667+00:00

Hi @Giden, Burcu

Have you replaced <resourcegroupname> and <subscriptionid> with your actual resource group and subscription details?
Giden, Burcu 20 Reputation points

2025-03-03T14:03:24.79+00:00

of course I am replcaing them. I didnt want to publish them here !!!

Accepted answer

0 additional answers

Your answer

Giden, Burcu 20 Reputation points

2025-03-01T10:35:25.7933333+00:00

Behavior After Force Unlink: After unlinking, is NorthEurope still active (as the primary), and will it serve all traffic until the failover completes?

Do I need to set up Azure Traffic Manager with a priority routing profile for proper failover handling between WestEurope and NorthEurope? Would this ensure proper routing when failover occurs, and traffic needs to be automatically directed to NorthEurope (secondary) during an outage in WestEurope?

Re-linking Issues: Can anyone guide me on how to correctly relink the Redis Enterprise clusters using Azure CLI or PowerShell? Is there a different approach to use for the re-linking process?
Giden, Burcu 20 Reputation points

2025-03-03T10:32:44.04+00:00

thanks for checking this.. however you have written exactly the same cli command that I have used to re-link them again- there is no difference in between... and it doesnt function..

the error I receive is "Failed to parse '--linked-databases' argument: "Model 'AAZObjectArg' has no field named 'resource'""
Giden, Burcu 20 Reputation points

2025-03-03T10:35:20.0133333+00:00

as written in my main question the az cli command is 1-1 the same with the one you wrote and it doesnt function !!! see the picture.. would appreciate any suggestions..
Sai Raghunadh M 2,880 Reputation points Microsoft External Staff

2025-03-03T14:01:15.8266667+00:00

Hi @Giden, Burcu

Have you replaced <resourcegroupname> and <subscriptionid> with your actual resource group and subscription details?
Giden, Burcu 20 Reputation points

2025-03-03T14:03:24.79+00:00

of course I am replcaing them. I didnt want to publish them here !!!

Answer 1

Hi @Giden, Burcu,

Behavior After Force Unlink:

NorthEurope still serves read/write operations, but no replication is happening. If WestEurope were unavailable, meaning no data sync between them during this period.

Re-linking Issues:

As per this MS document

Once the affected region's availability is restored, you need to delete the affected cache, and recreate it to add it back to your replication group.

When a region affected by an outage is restored, the Redis Enterprise cache in that region might still be in an inconsistent or degraded state. To ensure smooth operation and proper replication, you must delete and recreate the affected cache before adding it back to your replication group

Share via

Failover for Redis Enterprise with Active-geo-Replication

0 additional answers

Your answer