KSA Based Partners Solution
Google Could Platform offers a region in El Dammam in KSA, the region is called me-central2 and sold through CNTXT, GCP's exclusive reseller in the region.
Each zone is hosted in one or more clusters and Compute Engine independently maps zones to clusters for each organization.
As your provider in the Dammam region (me-central2) of Google Cloud Platform, ensuring your applications' uptime is our priority. Here's how we ensure disaster recovery:
Spread Across Zones: We deploy your services across multiple zones (me-central2-a, me-central2-b, me-central2-c) within the Dammam region. This way, if one zone experiences issues, your services remain unaffected in others.
Data Backup: We regularly back up your critical data to another region. In case of emergencies, your data is safe and can be quickly restored.
Traffic Distribution: Our load balancers efficiently distribute incoming traffic across all your instances and zones. If one zone gets busy or faces problems, traffic is automatically routed to others.
Monitoring and Alerts: We keep a constant watch on your infrastructure's health using monitoring tools. If anything goes wrong, we receive alerts immediately, allowing us to take prompt action.
Regular Testing: We conduct regular tests of our disaster recovery procedures to ensure they work when needed.
With these measures in place, your applications stay resilient, even in the face of unexpected challenges in the Dammam region.
We can use an existing Kubernetes cluster if the customer have one setup or install a new one, to install a new cluster we recommend the following configuration for high availability.
Make sure the cluster is Regional (HA) and select the me-central2 region (El Dammam region in KSA)
For the node pool configuration, we enable Autoscaler with a minimum of 1 node per region (3 total nodes) and a max of 3 nodes per region (9 in total)
For the nodes, our app workload require high memory to CPU ration so we recommend using e2-highmem-2 (2 vCPU, 1 core, 16 GB memory) nodes for the default node pool (used for DSquares apps) with 100 GB Balanced Persistent Disk.
In Cluster Security Tab, select enable workload identity (we use it to authenticate Kubernetes Service Accounts to GCP to give Kubernetes workloads access to google APIs ex: gcsfuse CSI addon)
In Cluster -> Features, make sure to check "Enable Cloud Storage FUSE CSI Driver" we use it as our RWX storage class CSI plugin
That's it for GKE Cluster, as for the DB:
Make sure you select SQL Server 2019, production preset setup, and "me-central2 (Dammam)" Region finally, make sure in "Zonal Availability" you select "Multiple zones (Highly available)"
Last updated