We Next Big Sound have been using an on-site MusicBrainz slave server for a couple of years now. We use the Web API to crawl the musicbrainz database and index data into our product. We are a financial supporter of the product.

We are using the new MusicBrainz docker platform found in the VM image located at ftp://vm.musicbrainz.org/pub/musicbrainz-vm/musicbrainz-server-2016-12-20.ova

Once the server is up we run 'bin/reset-containers' and 'bin/replicate now' to get the VM image up to the last replication packet that is valid for Schema 23.

At this point the Web API functions and we can perform our crawling.

We then perform the following to upgrade to 2017-05-15 schema update

• sudo docker exec -ti musicbrainzdocker_musicbrainz_1 bash
• git fetch origin && git checkout v-2017-05-15-schema-change
• ./upgrade.sh
• exit
• Set DB_SCHEMA_SEQUENCE to the NEW_SCHEMA_NUM in the DBDefs.pm

The upgrade and postgresql vacum seems to work BUT at this point the musicbrainz server no longer functions. Any queries to the web API return a 502 bad gateway.

Performing a bin/reset-containers does not fix anything.

I have also tried many of the suggestions at https://blog.musicbrainz.org/2017/05/16/schema-change-release-2017-05-15-including-upgrade-instructions/ but none of them seem to fix the issue.

At this point we are stuck at the last replication pre the schema change. Can you point us to instruction on how to go from the ova image to the lastest schema and replication packet.

If you need any logs from us please ask we will be happy to provide.

Thanks!