Rolling upgrades in Elasticsearch - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of rolling upgrade cluster health status

Consider a rolling upgrade process in Elasticsearch where nodes are upgraded one by one. After upgrading the first node, the cluster health is checked with the following command:

GET /_cluster/health

What is the expected value of the status field in the response during the upgrade?

Elasticsearch

GET /_cluster/health

A "yellow"

B "red"

C "green"

D "blue"

💡 Hint

During rolling upgrades, some shards may be temporarily unassigned.

During a rolling upgrade, some replicas may be temporarily unavailable while nodes are restarted. This causes the cluster health to be "yellow" because primary shards are available but some replicas are not yet assigned.

🧠 Conceptual

intermediate

Understanding shard allocation during rolling upgrades

During a rolling upgrade in Elasticsearch, what happens to shard allocation on the node that is being upgraded?

A Shards remain on the node until it is stopped, then replicas take over temporarily.

B All shards on the node are immediately moved to other nodes before upgrade starts.

C Shards are deleted and recreated after the node restarts.

D Shard allocation is frozen and no changes happen during the upgrade.

💡 Hint

Think about how Elasticsearch maintains data availability during node restarts.

Elasticsearch keeps shards on the node until it is stopped. When the node stops, replicas on other nodes serve the data to keep the cluster available during the upgrade.

❓ Predict Output

advanced

Result of incompatible index during rolling upgrade

You attempt a rolling upgrade from Elasticsearch 7.10 to 8.0. One index uses a deprecated mapping feature incompatible with 8.0. What will happen when the node with that index restarts?

A The node starts but the index is marked as read-only and inaccessible.

B The node fails to start and logs an error about the incompatible index.

C The node starts normally and automatically upgrades the index mapping.

D The node starts and deletes the incompatible index automatically.

💡 Hint

Elasticsearch enforces strict index compatibility during major version upgrades.

If an index uses deprecated or incompatible features, the node will fail to start and log errors until the index is fixed or deleted.

🔧 Debug

advanced

Identify the cause of cluster red status during rolling upgrade

During a rolling upgrade, the cluster health status unexpectedly turns red. Which of the following is the most likely cause?

A A master node was upgraded last, causing cluster state loss.

B The upgrade process skipped the cluster settings update step.

C A node was shut down before its shards were replicated elsewhere.

D All replicas were assigned to the same node, causing shard conflicts.

💡 Hint

Red status means some primary shards are missing or unassigned.

If a node is shut down before its shards are safely replicated, primary shards become unavailable, causing the cluster to turn red.

🚀 Application

expert

Planning a rolling upgrade with zero downtime

You need to perform a rolling upgrade on a 5-node Elasticsearch cluster with critical uptime requirements. Which sequence of steps ensures zero downtime?

A Disable shard allocation, upgrade all nodes simultaneously, then re-enable shard allocation.

B Upgrade master nodes first, then data nodes, restarting one node at a time with shard allocation enabled.

C Upgrade one node at a time with shard allocation enabled, waiting for cluster green status before next node.

D Disable shard allocation, upgrade one node at a time, wait for cluster green status before next node.

💡 Hint

Think about how shard allocation affects data availability during node restarts.

Disabling shard allocation before stopping each node for upgrade prevents unnecessary shard relocations, allowing faster recovery. After the node restarts and shards start, re-enabling allocation and waiting for green status ensures zero downtime.