Background:

In JIRA Data Center the search index will be copied between nodes. Example reasons for the copy include when a new node is added to the cluster, or when a node starts and detects a problem with its copy of the index.

This copy works by creating a backup zip file of the index and placing that file into the shared home. The node that needs the copy will then retrieve the copy and restore that copy.

Current State:

The logging that occurs does not allow an admin to follow this process.

• For example, there is no log message indicating that a node has requested a copy of the index from another node.
• On the other hand there is a log message on the node that creates the index backup.

Suggested Change:

Improve logging with the goal of giving JIRA admin clear indication that a node copy is happening and allowing that admin to follow the process

—

Examples of improvements:

• Add log entry when a node requests an index copy. Say that a copy is being requested and include reason the node is requesting the copy
  • The message could even give a brief description of what the next step is. "Requested index backup. Restore will occur after receiving 'Index Backed Up' message from another node"
• Improve logging during index backup process. Currently there are only two entries, "Index backup started" and "Index backup complete"
  • Suggestion: Add logging after each index is backed up. Similar to how the restore will log messages after certain steps are complete