cant get right partition lag from burrow

[Achillesxu]

burrow 1.3.6 cant get right partiton lag from kafka 0.10.2 burrow partition current lag metric is different from the lag from kafka-consumer-groups.sh --bootstrap-server localhost:9092 --describe --group logstash_for_es

[simonasr]

Might be related: we are seeing different LAG metric values after each Burrow service restart (produce/consume rates stays the same).

Blue annotations in the right graph indicates Burrow service restarts:

Burrow version: master branch @ 9bb078a3b7d4de21e66dbe150896a26ec9e4eadc Burrow config:

[general]
logdir="/var/log/burrow"
pidfile="/var/run/burrow/burrow.pid"
client-id="burrow-lagchecker"

[logging]
filename="/var/log/burrow/burrow.log"
level="info"
maxsize=100
maxbackups=30
maxage=10
use-localtime=false
use-compression=true

[zookeeper]
servers=["server_address:2181","another_server_address:2181"]
timeout=6
lock-path="burrow"

[httpserver.mylistener]
address=":9254"

[client-profile.logs-kafka]
client-id="burrow-logs-kafka"
kafka-version="2.4.0"

[cluster.logs-kafka]
class-name="kafka"
servers=["server_address:9092","another_server_address:9092"]
client-profile="logs-kafka"
topic-refresh=60
offset-refresh=15

[consumer.logs-kafka_consumers]
class-name="kafka"
start-latest=true
cluster="logs-kafka"
servers=["server_address:9092","another_server_address:9092"]