The kafka consumer group is used in reporting. For example, see consumer group lag grafana dashboard. The current random group that we use makes it difficult to know what the consumer really is. We should replace it by something more explicit.
Description
Details
Event Timeline
We still have to ensure in some way that it is distinct from others. We could however add hostname prefix or maybe make it somehow more clear to improve logging. Maybe use hostname + timestamp + random would make it better?
Change 416861 had a related patch set uploaded (by Smalyshev; owner: Smalyshev):
[wikidata/query/rdf@master] Change Kafka consumer ID to more recognizable string
Change 417115 had a related patch set uploaded (by Smalyshev; owner: Smalyshev):
[operations/puppet@production] Add consumer ID to Updater launch string
Change 417115 merged by Gehel:
[operations/puppet@production] Add consumer ID to Updater launch string
Mentioned in SAL (#wikimedia-operations) [2018-03-08T13:56:04Z] <gehel> restart wdqs-updater on wdqs1005 to validate new config option - T188716
Change 416861 merged by jenkins-bot:
[wikidata/query/rdf@master] Make Kafka consumer ID be command-line parameter
Change 418873 had a related patch set uploaded (by Gehel; owner: Smalyshev):
[operations/puppet@production] Add consumer ID to Updater launch string
Change 418873 merged by Gehel:
[operations/puppet@production] Add consumer ID to Updater launch string