Test build for docker

Expose max timestamp for each topic and/or partition (adobe#3)

* Expose max timestamp for each topic and/or partition


Kafka 3.0+ added support to fetch offset with the max timestamp.
This is useful in monitoring the steady progress of producing to a topic/partion
but also to detect very high timestamps in the topic that can stuck topic retention.

This patch adds two new metrics:
- kafka_topic_partition_max_timestamp: for each topic partition reports the max timestamp in that partition
- kafka_topic_max_timestamp: for each topic reports the max timestamp in that topic, computed as the maximum timestamp across partitions


See: 
- https://issues.apache.org/jira/browse/KAFKA-12541
- https://cwiki.apache.org/confluence/display/KAFKA/KIP-734%3A+Improve+AdminClient.listOffsets+to+return+timestamp+and+offset+for+the+record+with+the+largest+timestamp

* Update prometheus/collect_topic_partition_offsets.go

Set GOMAXPROCS automatically

This is required in containerised environments to properly set
GOMAXPROCS to actual container limits set and not the whole underlying node's CPUs

See:
https://github.com/uber-go/automaxprocs

Upgrade dependencies

Upgrade dependencies

Merge pull request adobe#2 from alexander-falca/empty-consumer-groups

Collect offsets for EMPTY groups

Merge pull request adobe#1 from cloudhut/master

Merge upstream changes

[INTERNAL] Don't report offsets and lag for empty consumer groups