chore(server): graceful shutdown

[iainsproat]

Description & motivation

When a deployment rolls out (causing a shutdown of an existing server), we see error messages from the Apollo server.

New connections were being made while the server was being shut down.

Apollo server was automatically shutting down when a SIGTERM or SIGINT signal was received; this created a race condition where it may close prior to the http server being drained of connections.

Changes:

• apollo server should not automatically close on receiving SIGTERM/SIGINT
• the apollo server should be shutdown after the server is drained of connections, this can be sequenced by the terminus package.
• the readiness endpoint should be managed by terminus, which ensures they respond unready if the server is being shutdown.
• the shutdown timeout (grace period) can be configured via environment variable

To-do before merge:

Screenshots:

Validation of changes:

Checklist:

• My pull request follows the guidelines in the Contributing guide?
• My pull request does not duplicate any other open Pull Requests for the same update/change?
• My commits are related to the pull request and do not amend unrelated code or documentation.
• My code follows a similar style to existing code.
• I have added appropriate tests.
• I have updated or added relevant documentation.

References

[gjedlicska]

this doesn't feel right to have as a module.

There is only application (express and apollo server) logic in here and no business logic.
I would put the code somewhere next to the app.ts file

[gjedlicska]

this logic should not be a module, its application logic