Custom ttl options for each collection and fixed tests

moscajs / aedes-persistence-mongodb

MongoDB persistence for Aedes

MIT License

14 stars 16 forks source link

Custom ttl options for each collection and fixed tests #26

Closed robertsLando closed 4 years ago

robertsLando commented 4 years ago

@mcollina

Changes

Feat: option to set custom TTL to each collection: subscriptions, incoming, outgoing, will, retained (keep back compatibility)
Feat: Set ttl to -1 to make a collection persistent (no ttl to docs)
Fix: Remove previous created indexes on connect if dropExistingIndexes option is specified
Added test for dropExistingIndexes
Fix: deprecation warning of url.parse with native-url npm package
Fix: test.js mongodb deprecation warnings
Updated mqemitter-mongodb to version 6.0.1
Feat Added createRetainedStreamCombi and outgoingEnqueueCombi methods
Dropped nodejs 6 support from .travis.yml
Added nodejs 12 support on .travis.yml
Docs: Updated readme with new features

robertsLando commented 4 years ago

@mcollina I have tried to upgrade aedes-persistence to 8.0.0 but there are some tests that fails. Seems all tests are related to the property messageId that isnull instead of undefined.

I don't know if this should be fixed in aedes-persistence tests abstract.js or in this library. Let me know if I need to change something.

mcollina commented 4 years ago

The behavior has changed, and the behavior of this library would have to be fixed.

robertsLando commented 4 years ago

Ok if you explain me what has changed I could try to fix it in another pr. What about this one? Can you make a review?

Daniel

On 13 Dec 2019, at 16:02, Matteo Collina notifications@github.com wrote:

The behavior has changed, and the behavior of this library would have to be fixed.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

mcollina commented 4 years ago

See https://github.com/mcollina/aedes-persistence/pull/31, https://github.com/mcollina/aedes-persistence/pull/33 and https://github.com/mcollina/aedes-persistence/pull/32

robertsLando commented 4 years ago

@mcollina Just one question: Why this very long timeout in the test?

robertsLando commented 4 years ago

@mcollina should be ok now

mcollina commented 4 years ago

I think dropping indexes at every start is going to be extremely disruptive for long-running operations. Can you keep that behavior behind an option?

robertsLando commented 4 years ago

@mcollina Yout right, fixed

robertsLando commented 4 years ago

@mcollina Should be ok now