Configuring vector connection to kafka

[k8s-comandante]

Hello everyone!

I have a Kafka cluster in my company that runs on the kerberos keytab, I want to use a vector agent in order to pick up microservices logs from Kafka, in the documentation https://vector.dev/docs/reference/configuration/sources/kafka/#:~:text=message_key-,librdkafka_options,-optional this case is hardly touched upon and hence there is a misunderstanding of how to implement it. Tried to configure sources like this but nothing happens:

[sources.kafka-logs.librdkafka_options]
"sasl.kerberos.keytab" = "/opt/file.keytab"

also tried https://component-pages--vector-project.netlify.app/docs/reference/sources/kafka/#:~:text=%23-,Advanced,-librdkafka_options.%22client.id

librdkafka_options. "sasl.kerberos.keytab" = "/opt/file.keytab"

but also unsuccessfully.

[jszwedko]

Hi @k8s-comandante ! Both of those should be correct ways to configure the rdkafka options. What makes you think they are not working?

[k8s-comandante]

Hey @jszwedko !
I specify a non-existent keytab file and the vector does not even say about it.

[jszwedko]

Ah I see. Looking at the configuration options page for librdkafka it seems like that value is only used in templating sasl.kerberos.kinit.cmd. Is it possible that there needs to be additional librdkafka configuration to use keberos? I see some additional documentation here: https://github.com/edenhill/librdkafka/wiki/Using-SASL-with-librdkafka that could be useful.