With introduction of settings, we are expecting the index dir path to have default.toml file

let result = Settings::new(config_dir_path);

The behaviour ACC to me should be

1. If path doesn't have default settings file, then copy from known location
2. If it has continue

Description

When infino server is running, tests fail due to a port conflict while starting rabbitmq.

Steps to Reproduce

1. Start infino server
2. Run cargo test -- --ignored

Expected behavior:

Tests should pass.

Actual behavior:

Tests fail due to a port conflict in rabbitmq.

Reproduces how often:

Every time.

Additional Information

Need to make sure the tests use a different rabbitmq port than the server.

What does this PR do?

• We try to connect to RabbitMq in exponential backoff fashion once docker is up and running

How does this PR work? (optional)

• Once docker containers are up, we are not certain when they'll start running so we should keep trying to connect to it

Refer to a related PR or issue link

• (Related PR or issue)

Description

error[E0599]: no function or associated item named new_with_max_params found for struct Tsldb in the current scope. With the change to include configuration, the method new_with_max_params was removed and changes were made to load from configuration, but the loom tests were not updated.

Steps to Reproduce

1. Run loom tests.

Expected behavior:

Tests should pass.

Actual behavior:

Tests are failing with error[E0599]: no function or associated item namednew_with_max_paramsfound for structTsldbin the current scope.

Additional Information

• NA

What does this PR do?

• This PR adds benchmark package to parent package

How does this PR work? (optional)

• Mentioned in README.md of benches folder

Refer to a related PR or issue link

• https://github.com/infinohq/infino/issues/10

Description

As of now, there isn't a way to cleanly showdown the server. We should provide a way to cleanly shutdown the running threads when Ctrl-C is pressed or SIGTERM is sent to the server process.

Additional Information

Write a test for clean server shutdown as part of this change.

Summary

Currently, infino always starts on port 3000. This needs to be configurable.

Motivation

Avoid port conflict in case the user has another service on port 3000.

Describe alternatives you've considered

NA

Additional context

NA

What does this PR do?

• Added issues and PR templates

Refer to a related PR or issue link

• #13

Checklist

• [X] I have written the necessary rustdoc comments.
• [X] I have added the necessary unit tests and integration tests. (for non-documentation changes)

• Use 'config' crate to create overridable configuration such as threshold number of log messages, rabbitmq container name, etc.
• The default configuration is specified in a default.toml file, and can be overridden using RUN_MODE specific toml file, and using environment variables.

To run

• Rust Fmt
• cargo check
• cargo test for infino and tsldb

Why?

These checks are done on every commit, so all the issues are caught before committing and code style is in check

A bunch of configuration (such as commit interval, max log messages per segment, and max data points per segment) is currently in code. This needs to be in a configuration managed in (yaml?) files with provision for overriding defaults on the command line and in tests.

Summary

Leverage GH workflows to:

• Run test when PR is raised.
• Run CI when pushed to main

Motivation

To make sure any build of main is error free, and expected functionalities work well.

Additional context

N.A.

Description

As of now, we only write to rabbitmq, but do not read from it at startup. We'll need to read messages from rabbitmq at startup, and apply them to the index. Also, periodically (at every commit?) we need to update rabbitmq with the 'marker' of which messages are applied to the index.

Additional Information

Write a test that Infino recovers without data loss in different crash scenarios.

Instead of iterating over each segment one by one, parallelly search the query in multiple segments.

We can use this https://docs.rs/rayon/latest/rayon/ to search concurrently.

Along with this PR, one should benchmark the optimal parameters for concurrency.