Skip to content

Database Lab Engine v3.2.0
Compare
Choose a tag to compare
@NikolayS NikolayS released this 19 Oct 17:51
· 461 commits to master since this release
v3.2.0
1ac2f12

Release v3.2.0 delivers a good number of new features and improvements. DLE UI is extended with two new tabs, "Logs" and "Configuration", to improve the user experience when configuring or reconfiguring DLE for managed Postgres services such as AWS RDS and Google Cloud SQL:

This allows users to see PGDATA initialization errors when setting up DLE, then adjust the configuration and retry initialization right in the UI – without any need to connect to the DLE machine using SSH.

The demo DLE instance can be found here: https://demo.aws.postgres.ai:446/instance (token: demo_token).

Postgres 15 is already supported!

PostgreSQL 15 is already supported by DLE and included in the list of default images, with many extensions shipped with it: https://postgres.ai/docs/database-lab/supported-databases.

AWS Marketplace – the easiest way to get DLE for any Postgres database

DLE in AWS Marketplace has now General Availability (GA) status, it has already been updated to version 3.2.0: https://aws.amazon.com/marketplace/pp/prodview-wlmm2satykuec, offering the Standard license, with support, embedded monitoring, simplified installation, and access to a wider variety of supported Postgres images (including extensions not available by default). An instance with DLE delivers "dev/test databases as a service" and levels up all dev&test processes. For example, on an r5.xlarge instance (4 vCPUs, 32 GiB RAM) with 200 GiB disk space, it is possible to run up to 30 database clones simultaneously (and even more, if shared_buffers is adjusted to lower values) spending as low as $360 per month – less than for a single regular RDS clone of the same size.

New contributor

🙇 The development team greets a new contributor: @Adrinlol. This release wouldn't be possible without other contributors: @agneum, @vitabaks, @DmitryFomin1, @ane4ka, @NikolayS, @denis256, @yuridevx.

What's new

New features

Improvements and fixes

  • Redesign the Retrieval service: !549
  • Mounting of socket directory for non-guest virtualization roles: !557 (@denis256)
  • Add a link to the guide "PostgreSQL configuration in clones": !559
  • Add session manager to CF template: !560
  • Add maintenance_work_mem for Logical mode: !567
  • Report the type of DLE subscription in UI: !563
  • Detect if DLE is running in a Docker container (this resolves issues of running DLE on KVM): !571, !603
  • Apply the configuration and then start Postgres on the restore stage, to reduce the number of restarts needed: !518, !547 (@yuridevx), !558, !572
  • Fix the creation of a restricted user: !586
  • Build CE UI Docker image after merging into the master branch: !590
  • Adjust container readiness health checks: !597, !602
  • Remove deprecated eslint analyzers: !605
  • Increase the number of parallel jobs in logical mode from 2/2 to 4/4 – usually, many more CPU cores are available: !611

Diff showing all changes

Diff between versions v3.1.2 and v3.2.0: v3.1.2...v3.2.0.

Migration from DLE v3.1.x and configuration notes

There are no special actions that are required to migrate from DLE 3.1.x to 3.2.0: you need to stop the DLE container, change the image, and start a new one using the new image. All snapshots should remain available, and if, at stopping time, some clones were present, they should be automatically recreated by DLE.

Note: if you have an older version, check the previous release notes as well: https://gitlab.com/postgres-ai/database-lab/-/releases.

Configuration directory

The config directory, by default, is ~/.dblab/engine/configs and the corresponding mounted volume must be writable:

  ...
  --volume ~/.dblab/engine/configs:/home/dblab/configs
  ...

Logs directory

Make sure that ~/.dblab/engine/logs is mounted as a volume to the DLE container:

  ...
  --volume ~/.dblab/engine/configs:/home/dblab/configs \
  --volume ~/.dblab/engine/meta:/home/dblab/meta \
  --volume ~/.dblab/engine/logs:/home/dblab/logs \
  ...

Alternatively, you can mount the entire directory (including configs and meta – but make sure that these subdirectories exist, DLE doesn't create them yet):

  ...
  --volume ~/.dblab/engine:/home/dblab
  ...

Query preprocessing

The queryPreprocessing section supports inline SQL now. Examples can be found here: !576#examples

Configuration UI

Because the configuration can be viewed and edited in the UI, make sure:

  • secret values are masked (e.g., verificationToken, Platform accessToken, you can find all masking rules here: !608#masking-rules)
  • DLE configuration matches the reference configuration files in structure and naming; it is not recommended to have custom names and sections

Where to get help

To get help, reach out to the Postgres.ai team and the growing community of Database Lab users and contributors: https://postgres.ai/contact.

Provide feedback, contribute

Any feedback will be greatly appreciated:

If you want to help the project grow, consider various ways of contributing:

Contributors

ane4ka, NikolayS, and 6 other contributors
Assets 2
Loading
0 Join discussion