Releases and Migrations
We are continuously working on new features, catching bugs and improving our architecture. To get the latest changes to you faster, our pipeline produces releases quite often. You can always find the latest version on GitHub. On this page we provide all the necessary knowledge you need to always keep DIVA up to date.
We recommend that you always use the latest stable version of DIVA. The latest version of the source code can always be found in the main branch. The latest unstable development state can be found in dev branch.
In most cases you can safely use the newer patch or minor version of DIVA. However, it is possible that we may change the internal infrastructure or add new features. Then the existing data in your DIVA instance should also be brought up to date. Most migrations are mandatory.
The update of the major version indicates a breaking change and the migration is necessary to guarantee the operability of the system.
Releases | Changelog | Migration tool
Release | Migration |
---|---|
✅️ Not required | |
🆘 Mandatory | |
✅️ Not required | |
⚠️ Recommended | |
✅️ Not required | |
✅️ Not required | |
✅️ Not required | |
❕ Not available |
Upgrade to the latest version
To bring DIVA up to date, simply download the latest Docker distribution docker-deployment.zip
from the release page.
In both cases navigate to docker/
directory to execute the start script:
./up_core.sh
Migration execution
To prepare for the migration, deploy the target DIVA version and then execute the migration, if available. Make sure your DIVA instance is not actively used in a production environment until the migration is completed to avoid potential inconsistencies.
We ship DIVA with a migration Docker service that you can also find in the docker/
directory.
This service is attached to two specific versions, for which the migration is intended.
Since not all releases need a migration, you may not find a migration service for a particular version.
If you want to upgrade from an older version, you must run all intermediate migrations sequentially.
For example, to upgrade from v3.0.3
to 3.2.0
, you have to deploy 3.2.0
and execute the migration for 3.1.0
.
Execute the migration service with the following:
./migrate.sh
caution
Make a snapshot of your DIVA environment. It should be ok to save the volumes.
caution
After a successful execution, the service should not be executed again, as this may cause undesired side effects.
Releases
v3.3.1
No migration from previous version required
v3.3.0
A mandatory migration is available for DIVA 3.3.0. We provide a migration service to upgrade DIVA 3.2.0 to 3.3.0.
Affected Components:
- JSON Schema:
licenses
now is an object with more fields. Therefore we need to migrate thelicenses
field in ourdocumentstore
MongoDB and Elasticsearch.
v3.2.0
No migration from previous version required
v3.1.0
The recommended migration is available for DIVA 3.1.0. You can use the migration service to upgrade from DIVA >=3.0.0.
Affected Components:
DIVA Lake Adapter (DLA) - create dedicated MongoDB database, create
id
to uploaded filehash
mapping. Required for new Resources delete functionalityDCS Adapter (DA) - create dedicated MongoDB database, migrate MongoDB old
dsc
collection to new database and creatid
tooffer
mapping for all on DSC offered resources. Required for new Resources delete functionality
v3.0.3
No migration from previous version required
v3.0.2
No migration from previous version required
v3.0.1
No migration from previous version required
v3.0.0
No migration service available. Manual migration required