docs: refine option one as the preferred one

Author: ctron

docs/adrs/00008-re-process-documents.md

@@ -14,8 +14,8 @@ in the database. We also store the original source document "as-is".
 When making changes to the database structure, we also have a migration process, which takes care of upgrading the
 database structures during an upgrade.
 
-However, in some cases, changing the database structure actually means to extract more information from documents and is
-currently stored in the database. Or information is extracted in a different way. This requires a re-processing of
+However, in some cases, changing the database structure actually means extracting more information from documents than
+is currently stored in the database. Or information is extracted in a different way. This requires a re-processing of
 all documents affected by this change.
 
 ### Example
@@ -30,25 +30,39 @@ This ADR makes the following assumptions:
 * All documents are stored in the storage
 * It is expected that an upgrade is actually required
 * Running such migrations is expected to take a long time
+* The management of infrastructure (PostgreSQL) is not in the scope of Trustify
 
 Question? Do we want to support downgrades?
 
 ## Decision
 
-### Option 1
+During the migration of database structures (sea orm), we also re-process all documents (if required). This would
+be running during the migration job of the Helm chart and would have an impact on updates as the rollout of newer
+version pods would be delayed until the migration (of data) has been finished.
 
-During the migration of database structures (sea orm), we also re-process all documents (when required).
+This would also require to prevent users from creating new documents during that time. Otherwise, we would need to
+re-process documents ingested during the migration time. A way of doing this could be to leverage PostgreSQL's ability
+to switch into read-only mode. Having mutable operations fail with a 503 (Service Unavailable) error. This would also
+allow for easy A/B (green/blue) database setups. Switching the main one to read-only, having the other one run the
+migration.
 
-In order to report progress, we could write that state into a table and expose that information to the user via the UI.
+We could provide an endpoint to the UI, reporting the fact that the system is in read-only mode during a migration.
 
-* 👎 Might serve inaccurate data for a while
-* 👎 Might block an upgrade if re-processing fails
 * 👍 Can fully migrate database (create mandatory field as optional -> re-process -> make mandatory)
+* 👍 Might allow for an out-of-band migration of data, before running the upgrade (even on a staging env)
+* 👍 Would allow to continue serving data while the process is running
 * 👎 Might be tricky to create a combined re-processing of multiple ones
+* 👎 Might block an upgrade if re-processing fails
+
+## Open items
+
+* [ ] How to handle unparsable or failing documents during migration?
+
+## Alternative approaches
 
 ### Option 2
 
-We create a similar module as for the importer. Running migrations after an upgrade. Accepting that in the meantime, 
+We create a similar module as for the importer. Running migrations after an upgrade. Accepting that in the meantime,
 we might service inaccurate data.
 
 * 👎 Might serve inaccurate data for a while for a longer time
@@ -58,7 +72,7 @@ we might service inaccurate data.
 
 ### Option 3
 
-We change ingestion in a way to it is possible to just re-ingest every document. Meaning, we re-ingest from the 
+We change ingestion in a way to it is possible to just re-ingest every document. Meaning, we re-ingest from the
 original sources.
 
 * 👎 Might serve inaccurate data for a while for a longer time
@@ -68,14 +82,10 @@ original sources.
 * 👎 Won't work for manual (API) uploads
 * 👎 Would require removing optimizations for existing documents
 
-## Open items
-
-…
-
-## Alternative approaches
-
-…
 
 ## Consequences
 
-…
+* The migration will block the upgrade process until it is finished
+* Ansible and the operator will need to handle this as well
+* The system will become read-only during a migration
+* The UI needs to provide a page for monitoring the migration state. The backend needs to provide appropriate APIs.