[ https://issues.apache.org/jira/browse/KAFKA-4108?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15491082#comment-15491082 ]
Vahid Hashemian commented on KAFKA-4108: ---------------------------------------- [~hachikuji] Would changing the output format from {code} key: metadata::cgroup1 payload: consumer:range:1:{consumer-1-040134f0-4d2a-4665-b996-a59322cc82cf=[t1-0]} {code} to {code} key: {metadata:cgroup1} payload: {protocolType: consumer, groupMetadata.protocol:range, groupMetadata.generationId:1, assignment:{consumer-1-040134f0-4d2a-4665-b996-a59322cc82cf=[t1-0]}} {code} be sufficient for this? > Improve DumpLogSegments offsets-decoder output format > ----------------------------------------------------- > > Key: KAFKA-4108 > URL: https://issues.apache.org/jira/browse/KAFKA-4108 > Project: Kafka > Issue Type: Improvement > Components: tools > Reporter: Jason Gustafson > Assignee: Vahid Hashemian > > When using the DumpLogSegments with the "--offsets-decoder" option (for > consuming __consumer_offsets), the encoding of group metadata makes it a > little difficult to identify individual fields. In particular, we use the > following formatted string for group metadata: > {code} > ${protocolType}:${groupMetadata.protocol}:${groupMetadata.generationId}:${assignment} > {code} > Keys have a similar formatting. Most users are probably not going to know > which field is which based only on the output, so it would be helpful to > include field names. Maybe we could just output a JSON object? -- This message was sent by Atlassian JIRA (v6.3.4#6332)