Docker ES 5.5.0 Low performance when ingest 1000 docs/sec


(Rodrigo Nicolas Garcia) #1

Hello,
First the general data:

  • Version ES: 5.5.0
  • Version ingestion: logstash-input-kafka (4.2.0)
  • Operating System: docker - official compiled docker file:
FROM docker.elastic.co/logstash/logstash:5.5.0

RUN rm -f /usr/share/logstash/pipeline/logstash.conf \
    && rm -f /usr/share/logstash/config/logstash.yml \
    && rm -f /usr/share/logstash/config/log4j2.properties

ADD pipeline/ /usr/share/logstash/pipeline/
ADD config/ /usr/share/logstash/config/

RUN chown -R logstash:logstash /usr/share/logstash/config/ \
    && chmod -R +x /usr/share/logstash/config/ \
    && logstash-plugin remove logstash-input-kafka \
    && logstash-plugin remove logstash-output-kafka \
    && logstash-plugin install --version 4.2.0 logstash-input-kafka \
    && logstash-plugin install --version 4.0.4 logstash-output-kafka \
    && logstash-plugin list --verbose kafka

Config File (if you have sensitive info, please remove it): Example input config:

bootstrap_servers => "${KAFKA_BOOTSTRAP_RC}"
client_id => "${HOSTNAME}_rc"
group_id => "${CONSUMER_GROUP}_rc"
topics_pattern => "${KAFKA_TOPICS}"
consumer_threads => "12"
decorate_events => true
auto_offset_reset => "earliest"
heartbeat_interval_ms => "300"
fetch_max_wait_ms => "300"
fetch_min_bytes => "20971520"
poll_timeout_ms => 300

Logstash Workers : 12 for each 6 nodes.
Question:
I would like consume 2000 messages/sec. But only consume 200/sec per node whit this config.
Also when they ingest ES get some bad performance.
I have 5 instances (all master/ingest/data)
Gretings.

Edit: add some images:
logstash performance:


One node perfermance:

Elastic overview:

Elastic Nodes general:

Master performance:

i don't know for what see that gaps.


(system) #2

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.