tl;dr

Dry run track

- Include the last entity + solving the long run issue

- Have a last dry run

Product adaptations track

- Finish the UID adaptations

- Last check to get ready for the migration

Search track

- deliver the two lambdas for search and indexing

More details below 👇

Product adaptations

We are in the process of integrating a new database into the different systems of our product. The main idea is to make them able to communicate with the legacy and the new db simultaneously. We can monitor the new database behaviour using the legacy one as a source of truth. This is what we call the shadowing process. It will allow us to run the new system without taking the risk of a major incident involving data corruption in case of an issue.

Once done, we will switch to the new database.

We can't activate the shadowing process before having all sub-systems manage the new database. However, we can release progressively to ensure non-regression.

All product adaptations should be completed this week. We will also focus on preparing for

the shadowing process

Integrating a new search

We defined a strategy to implement a new search engine a few weeks ago, going through intermediate steps. We have decided to make a bet and work directly towards the target search engine.

The team has started implementing the new search with OpenSearch. There will be three major steps:

- Implementing a lambda in charge of indexing documents

- Implementing a lambda in charge of searching for documents

- Implementing a JS client so the Page Builder can use the search capabilities

We are expecting to complete the first two steps this week.

Dry run

This week, we are completing the dry run for

interesting versions,

At this stage, the dry-run process lasts less than 3 hours for the 20k most active repos.