search

openzipkin / openzipkin.github.io

content for https://zipkin.io
https://zipkin.io
Apache License 2.0
39 stars 61 forks source link

Clarify state of storages #161

Open jcchavezs opened 3 years ago

jcchavezs commented 3 years ago

From https://zipkin.io/pages/extensions_choices.html

Zipkin server bundles extension for span collection and storage. By default spans can be collected over http, Kafka or RabbitMQ transports and stored in-memory or in MySQL, Cassandra or Elasticsearch.

the storage part is quite vague, we should either put a link to each of them or add hints around the state of at least those who aren't meant to be used in prod (e.g. in memory and mysql)

Related: https://gitter.im/openzipkin/zipkin?at=5fae5015c10273610a020a11

codefromthecrypt commented 3 years ago

Vague might not be the best word as we literally say what is available (though missing that activemq is also there).

Want to copy/pasta in the missing ActiveMQ, and then add another sentence. "Production sites should only consider Cassandra or Elasticsearch storage"?

jcchavezs commented 3 years ago

Vague might not be the best word as we literally say what is available (though missing that activemq is also there). I am sorry, not a native here.

Shall we mention kafka or at least point out a section of contrib filtered by storage?

José Carlos Chávez

fre. 13. nov. 2020 kl. 11:43 skrev Adrian Cole notifications@github.com:

Vague might not be the best word as we literally say what is available (though missing that activemq is also there).

Want to copy/pasta in the missing ActiveMQ, and then add another sentence. "Production sites should only consider Cassandra or Elasticsearch storage"?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/openzipkin/openzipkin.github.io/issues/161#issuecomment-726692167, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXOYAQFCTGBH2P6FWIDLLLSPUEWJANCNFSM4TULS37Q .

codefromthecrypt commented 3 years ago

This feels easier to do in a PR, but taking a stab:

Zipkin server bundles extension for span collection and storage. By default spans can be collected over ActiveMQ, gRPC, HTTP, Kafka, RabbitMQ and Scribe transports and stored in-memory or in MySQL, Cassandra or Elasticsearch. Production sites should only consider Cassandra or Elasticsearch storage.

...

There's already a section at the bottom with Kafka storage, but agree it is amongst dead projects also not in our org. How about dropping the dead projects, retaining storage-kafka and renaming the section to "Contrib"?