GitHub user julienlagorsse-loreal edited a comment on the discussion:
ElasticSearch Logs not working
We have the same bug, connection seems to be fine in webserver as we get 200 on
some count request however we get this in webserver log page
*** Log
debug_airflow_to_es-print_debug_message-manual__2024-12-06T12:23:30.109158+00:00--1-1
not found in Elasticsearch. If your task started recently, please wait a
moment and reload this page. Otherwise, the logs for this task instance may
have been removed.
When I check on elasticsearch there is no index.
The triggerer seems to create a wrapper to redirect to stdout to elasticsearch,
we don't really understand the NOTSET part
2024-12-06 13:26:10 [2024-12-06T12:26:10.912+0000]
{triggerer_job_runner.py:181} INFO - Setting up TriggererHandlerWrapper with
handler <ElasticsearchTaskHandler (NOTSET)>
2024-12-06 13:26:10 [2024-12-06T12:26:10.914+0000]
{triggerer_job_runner.py:237} INFO - Setting up logging queue listener with
handlers [<RedirectStdHandler <stdout> (NOTSET)>, <TriggererHandlerWrapper
(NOTSET)>]
2024-12-06 13:26:10 [2024-12-06T12:26:10.920+0000]
{triggerer_job_runner.py:338} INFO - Starting the triggerer
We don't really use this triggerer for now so it doesn't help much
We can see Listening at: http://[::]:8794 for triggerer
For Worker we don't see a wrapper logs, but we see the listening
http://[::]:8793
but nothing else seems to be really related to logging.
I don't find new logs on /opt/airflow/logs, scheduler logs are there but
nothing special in it (I have set a cron every min to my test dag)
We aren't really if it's the responsibility of the worker to push logs on
elasticsearch, and we don't know who is responsible to write logs.
Elasticsearch is hosted in a docker compose, same network, curl
elasticsearch:9200 works everywhere (worker, webserver, ...)
AIRFLOW_CONN_ELASTICSEARCH_DEFAULT: "http://elasticsearch:9200";
_PIP_ADDITIONAL_REQUIREMENTS:
${_PIP_ADDITIONAL_REQUIREMENTS:-apache-airflow-providers-elasticsearch}
[logging]
logging_level = INFO
remote_logging = True
remote_log_conn_id=elasticsearch_default
[elasticsearch]
host = http://elasticsearch:9200
write_stdout = True
json_format = True
[elasticsearch_configs]
verify_certs=False
GitHub link:
https://github.com/apache/airflow/discussions/23734#discussioncomment-11484605
----
This is an automatically sent email for [email protected].
To unsubscribe, please send an email to: [email protected]
