Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Log files are stored as ${sys:druid.node.type}.log #15967

Closed
Subhashini2610 opened this issue Feb 26, 2024 · 5 comments · Fixed by #16282
Closed

Log files are stored as ${sys:druid.node.type}.log #15967

Subhashini2610 opened this issue Feb 26, 2024 · 5 comments · Fixed by #16282

Comments

@Subhashini2610
Copy link

I am using a k8s based deployment of Druid and see that each node creates log files that are named as
${sys:druid.node.type}.log

The variable is not getting resolved to the right node type. The log files are stored as ${sys:druid.node.type}.log or ${sys:druid.node.type}-<date>.log

Affected Version

Druid version: 28.0.0

Description

Due to the above problem, I am not able to index the logs appropriately. Am I missing something here?

@Subhashini2610
Copy link
Author

Please help here

@abhishekagarwal87
Copy link
Contributor

when you log into the pod, can you check the full command line that was used to start the druid process?

@Subhashini2610
Copy link
Author

@abhishekagarwal87 How do I check that? The logs of the broker only contain the following:

2024-02-26T14:50:22+00:00 startup service broker
Setting druid.host=10.0.2.188 in /tmp/conf/druid/cluster/query/broker/runtime.properties
Setting druid.auth.authenticator.BasicMetadataAuthenticator.skipOnFailure=false in /tmp/conf/druid/cluster/query/broker/runtime.properties
Setting druid.broker.http.numThreads=15 in /tmp/conf/druid/cluster/query/broker/runtime.properties
Setting druid.indexer.logs.type=file in /tmp/conf/druid/cluster/query/broker/runtime.properties
Setting druid.emitter.statsd.port=8125 in /tmp/conf/druid/cluster/query/broker/runtime.properties
...
...
...

Also, the below is the broker container spec in my helm chart.

Screenshot 2024-02-26 at 9 14 06 PM

All other nodes also have similar spec, except the args, which refer to the particular node type like coordinator, historical etc.

@fectrain
Copy link
Contributor

fectrain commented Apr 13, 2024

Hi, @Subhashini2610 , I met the similar issue.
Actually I'm using Helm-druid to manage the K8s deployment.
Usually the different services are deployed in different pods, So I'll just list out all pods(kubectl get pods -n <namespace>) and then ssh to the pod with specific pod_name, by this way I'm able to identify the which service/pod I'm accessing, even though the name of log_file looks like ${sys:druid.node.type}.log

@fectrain
Copy link
Contributor

fectrain commented Apr 15, 2024

Hi, @Subhashini2610 @abhishekagarwal87 , the issue is probably due to druid.node.type not been specified when starting up, whose value is supposed to be the service type(e.g. broker/historical, etc..), have just made some fix right here

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants