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

Need more doc about upgrading from 2.2 to 2.3 #5568

Closed
dedemorton opened this issue Jun 29, 2016 · 4 comments
Closed

Need more doc about upgrading from 2.2 to 2.3 #5568

dedemorton opened this issue Jun 29, 2016 · 4 comments
Assignees
Labels

Comments

@dedemorton
Copy link
Contributor

The current upgrade documentation needs to discuss how the change to the default output workers that we made in 2.3 (see issue #4877) might affect customers who are upgrading from 2.2 to 2.3.

We need to:

  • Update the Breaking Changes doc to indicate that the change to output workers is potentially a breaking change that may affect throughput.
  • Also mention in the Release Notes that this change might affect throughput.
  • Add a topic about upgrading to 2.3 and mention that in order to get the same configuration in 2.3 as you had in 2.2, you will need to specify the number of workers in the actual output configuration (note the value specified for output workers depends on whether they were using -w.
  • In the plugin docs, it would also be good to remove or reword the following sentence that appears under the description for workers because it's so vague that it's virtually meaningless: "Note that this setting may not be useful for all outputs."
@dedemorton dedemorton self-assigned this Jun 29, 2016
@GlenRSmith
Copy link
Contributor

👍

@suyograo
Copy link
Contributor

Add a topic about upgrading to 2.3 and mention that in order to get the same configuration in 2.3 as you had in 2.2, you will need to specify the number of workers in the actual output configuration (note the value specified for output workers depends on whether they were using -w.

I think this statement needs to be qualified to say output workers will increase resource usage. See explanation below:

Its a bug that in 2.2 we made output workers = pipeline workers = # of cores. Its perfectly fine to set pipeline workers = # of cores or even greater than # of cores. Outputs in Logstash are not thread safe in terms of underlying resources it uses. Increasing output workers (for Kafka, Elasticsearch) may increase the throughput but it will also result in resource issues like too many file handles being used, too many sockets being opened etc. For reference see #4754. This has also been pointed out in the blog: https://www.elastic.co/blog/upgrade-guide-for-logstash-2-2. See section "New Pipeline and Outputs".

@dedemorton
Copy link
Contributor Author

Wanted to note that changing the value of output workers (the fix described above) is not generally recommended. The dev team is working on a better explanation to help users understand and resolve this problem if they encounter it when moving to 2.3.

@dedemorton
Copy link
Contributor Author

Closing because we decided to address the concerns that prompted this issue by adding a troubleshooting guide instead.

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

No branches or pull requests

3 participants