After upgrading LDES Server, new ingested LDES member restarts the sequence from 1

[xdxxxdx]

Prerequisites

• ldes/ldes-server:1.4.0-SNAPSHOT
• Test directory: https://github.com/Informatievlaanderen/VSDS-LDES-E2E-testing/tree/main/tests/021.server-upgrade
• Old LDES Server: LDES_OLD_SERVER_TAG:-20230405T0843 is up and running
• 314 LDES members are ingested to the LDES Server

To Reproduce

1. Pause the workbench output. curl -X POST "http://localhost:8081/admin/api/v1/pipeline/halt"
2. Bring old server down (stop it, remove volumes and image without confirmation). docker compose rm --stop --force --volumes old-ldes-server
3. Launch new server, wait until database migrated and server started docker compose up new-ldes-server -d docker logs --tail 1000 -f $(docker ps -q --filter "name=new-ldes-server$")
4. Resume the workbench output. curl -X POST "http://localhost:8081/admin/api/v1/pipeline/resume"

Current behavior 851 NEW LDES members are ingested to the new LDES Server, together with the OLD 315 LDES members. => 1165 LDES members are in the iow_devices.ingest_ldesmember table The end of the sequenceNr is 851 instead of 1165

The sequenceNr is restarted from 1 from the new ingest members of the new LDES Server

Table dump: iow_devices.ingest_ldesmember.zip

Expected behavior All 1165 ldes members should be in a continuously sequence?. e.g. from 1 to 1165

[Yalz]

Hi @xdxxxdx , Although this issue is still there, the reality is that this has very little impact on current customers. Everyone who used a ldes-server older than ~june 2023 might have this behaviour. But the worst thing that could come of this, is that a few members are fragmented out of order.

However since we don't have any more onboarders who haven't migrated yet. This is to be a non issue