SUSE Observability Kubernetes Deployment: Non-HA vs. HA Profiles

The SUSE Observability platform supports two deployment profiles: Non-High Availability (Non-HA) and High Availability (HA). These profiles determine the level of fault tolerance, scalability, and resource distribution across the nodes. Below is a detailed comparison of the pod configurations for each profile:

To gain a better understanding of each pod function in SUSE Observability, please review the following link: Overview of Subsystems.

Non-High Availability (Non-HA)

In a Non-HA setup, all services are designed to run as single instances, making the deployment lightweight and suitable for environments where high availability is not a critical requirement. Key characteristics include:

• Single instance of core components such as Elasticsearch, Kafka, Zookeeper, and Clickhouse.
• Minimal redundancy across the system, simplifying resource allocation and management.
• Example of pods in a Non-HA setup:
  • suse-observability-clickhouse-shard0-0
  • suse-observability-elasticsearch-master-0
  • suse-observability-kafka-0
  • suse-observability-zookeeper-0
  • suse-observability-ui-xxx

Click to expand pods output in Non-HA setup

NAME                                                              READY   STATUS    RESTARTS        AGE
suse-observability-clickhouse-shard0-0                            2/2     Running   0               2d22h
suse-observability-correlate-d5dc56d8b-5t6lf                      1/1     Running   0               2d18h
suse-observability-e2es-754f9c858-sjfz6                           1/1     Running   0               2d22h
suse-observability-elasticsearch-master-0                         1/1     Running   0               2d22h
suse-observability-hbase-stackgraph-0                             1/1     Running   0               2d22h
suse-observability-hbase-tephra-0                                 1/1     Running   0               2d22h
suse-observability-kafka-0                                        2/2     Running   0               2d22h
suse-observability-kafkaup-operator-kafkaup-7c57c5bdbf-zksfr      1/1     Running   0               2d22h
suse-observability-otel-collector-0                               1/1     Running   0               2d22h
suse-observability-prometheus-elasticsearch-exporter-765cd6w5sl   1/1     Running   0               2d22h
suse-observability-receiver-759754bb48-ttbg7                      1/1     Running   0               2d18h
suse-observability-router-7cbfd778d9-hkckv                        1/1     Running   0               2d22h
suse-observability-server-86f4649d4b-r82tp                        1/1     Running   0               2d18h
suse-observability-ui-56c7df6cd7-qlss2                            2/2     Running   0               2d22h
suse-observability-victoria-metrics-0-0                           1/1     Running   0               2d22h
suse-observability-vmagent-0                                      1/1     Running   0               2d22h
suse-observability-zookeeper-0                                    1/1     Running   0               2d22h

High Availability (HA)

In an HA deployment, the architecture introduces redundancy and scalability to ensure continuous service availability and handle larger workloads. The HA profile includes multiple instances of critical components, enabling the system to withstand node or pod failures. Key features include:

• Elasticsearch: Operates with three master nodes (elasticsearch-master-0, elasticsearch-master-1, elasticsearch-master-2) to maintain quorum.
• Kafka and Zookeeper: Three replicas each, ensuring high throughput and resiliency.
• HBase and HDFS: Includes multiple region servers, data nodes, and master nodes to distribute data and maintain uptime.
• suse-observability-server-xxxx will be replaced by suse-observability-notification-xxx,suse-observability-slicing-xxx and suse-observability-initializer-xxx in HA setup.
• ** Other components are in HA, such as:
  • suse-observability-health-sync
  • suse-observability-receiver-logs, suse-observability-receiver-process-agent, and suse-observability-receiver-base for better load segregation.

Example pods in HA:

• suse-observability-elasticsearch-master-[0, 1, 2]
• suse-observability-kafka-[0, 1, 2]
• suse-observability-zookeeper-[0, 1, 2]
• suse-observability-hbase-hbase-rs-[0, 1, 2]
• suse-observability-ui-[multiple instances]

Click to expand pods output in HA setup

NAME                                                              READY   STATUS    RESTARTS      AGE
suse-observability-agent-checks-agent-95ff66f4c-l974n             1/1     Running   0             10m
suse-observability-agent-cluster-agent-7b85bc7857-kv5fw           1/1     Running   0             10m
suse-observability-agent-logs-agent-4ppw7                         1/1     Running   0             10m
suse-observability-agent-logs-agent-x9jn8                         1/1     Running   0             10m
suse-observability-agent-node-agent-8nf87                         2/2     Running   0             10m
suse-observability-agent-node-agent-gvqdp                         2/2     Running   0             10m
suse-observability-api-7d796dcf9b-qkf9q                           1/1     Running   0             36m
suse-observability-checks-589488bd64-sfw6j                        1/1     Running   0             15m
suse-observability-clickhouse-shard0-0                            2/2     Running   0             35m
suse-observability-correlate-5cbbb98d46-pwdx6                     1/1     Running   0             33m
suse-observability-correlate-5cbbb98d46-rjfl5                     1/1     Running   0             36m
suse-observability-correlate-5cbbb98d46-tcpn6                     1/1     Running   0             32m
suse-observability-e2es-7649f4fdb4-st4rx                          1/1     Running   0             36m
suse-observability-elasticsearch-master-0                         1/1     Running   0             15m
suse-observability-elasticsearch-master-1                         1/1     Running   0             15m
suse-observability-elasticsearch-master-2                         1/1     Running   0             15m
suse-observability-hbase-hbase-master-0                           1/1     Running   0             34m
suse-observability-hbase-hbase-master-1                           1/1     Running   0             35m
suse-observability-hbase-hbase-rs-0                               1/1     Running   0             34m
suse-observability-hbase-hbase-rs-1                               1/1     Running   0             34m
suse-observability-hbase-hbase-rs-2                               1/1     Running   0             35m
suse-observability-hbase-hdfs-dn-0                                1/1     Running   0             32m
suse-observability-hbase-hdfs-dn-1                                1/1     Running   0             33m
suse-observability-hbase-hdfs-dn-2                                1/1     Running   0             35m
suse-observability-hbase-hdfs-nn-0                                1/1     Running   0             34m
suse-observability-hbase-hdfs-snn-0                               1/1     Running   0             34m
suse-observability-hbase-tephra-0                                 1/1     Running   0             34m
suse-observability-hbase-tephra-1                                 1/1     Running   0             35m
suse-observability-health-sync-8cb88d7d-c62j4                     1/1     Running   0             15m
suse-observability-initializer-7d487dffb7-l858s                   1/1     Running   0             35m
suse-observability-kafka-0                                        2/2     Running   0             33m
suse-observability-kafka-1                                        2/2     Running   0             34m
suse-observability-kafka-2                                        2/2     Running   0             35m
suse-observability-kafkaup-operator-kafkaup-7c57c5bdbf-j8hlr      1/1     Running   0             43m
suse-observability-notification-5d999c4d5b-4j8s5                  1/1     Running   0             35m
suse-observability-otel-collector-0                               1/1     Running   0             43m
suse-observability-prometheus-elasticsearch-exporter-54cc6lbhf5   1/1     Running   0             43m
suse-observability-receiver-base-54b66d775d-q6ftf                 1/1     Running   0             36m
suse-observability-receiver-logs-6d557fd5f5-bv4tt                 1/1     Running   0             36m
suse-observability-receiver-process-agent-7d4f8bc7c5-7bc7m        1/1     Running   0             36m
suse-observability-router-bd8d7b544-68dr8                         1/1     Running   0             43m
suse-observability-slicing-74bfddd9b9-wnh6q                       1/1     Running   0             35m
suse-observability-state-5d9bd7f694-ddzzl                         1/1     Running   0             36m
suse-observability-sync-7fbb946b4b-mddrm                          1/1     Running   0             36m
suse-observability-ui-7c548d877c-48ndx                            2/2     Running   0             43m
suse-observability-ui-7c548d877c-7ss4g                            2/2     Running   0             43m
suse-observability-victoria-metrics-0-0                           1/1     Running   0             35m
suse-observability-victoria-metrics-1-0                           1/1     Running   0             21m
suse-observability-vmagent-0                                      1/1     Running   0             22m
suse-observability-zookeeper-0                                    1/1     Running   0             35m
suse-observability-zookeeper-1                                    1/1     Running   0             35m
suse-observability-zookeeper-2                                    1/1     Running   0             35m

Comparison Summary

Feature	Non-HA Deployment	HA Deployment
Scalability	Limited	Highly scalable
Redundancy	None	Multiple replicas for core services
Resilience to Failures	Low	High
Resource Requirements	Minimal	High
Example Pods	kafka-0, zookeeper-0	kafka-[0, 1, 2], zookeeper-[0, 1, 2]

If you're planning to deploy SUSE Observability in a production environment, the HA setup is strongly recommended. It provides the necessary redundancy and fault tolerance to ensure that your observability platform remains operational even in the face of failures.