V10 Beta is ready! 📦

We just released the beta version of the V10 release :100: . This contains many quality-of-life tweaks to the front-end (vertical navbar, session expiry won’t lose a user’s place in the app), improved visualisations (fixed table visualisation header, horizontal bar charts) and stability upgrades (typescript, Antd v4). For a full list of changes, we encourage you to checkout the CHANGELOG.

As always this release wouldn’t be possible without the help of our community. This version had contributions from over 35 people :pray:

Our special thanks to Sohail Ahmed for disclosing a vulnerability in our “forgot password” page which we patched in #5425.

Upgrading

(This section is duplicated from the previous release - since many users will upgrade directly from V8 → V10)

Typically, if you are running your own instance of Redash and wish to upgrade, you would simply modify the Docker tag in your docker-compose.yml file. Since RQ has replaced Celery in this version, there are a couple extra modifications that need to be done in your docker-compose.yml :

  1. Under services/scheduler/environment , omit QUEUES and WORKERS_COUNT (and omit environment altogether if it is empty).
  2. Under services , add a new service for general RQ jobs:

worker: <<: *redash-service command: worker environment: QUEUES: “periodic emails default” WORKERS_COUNT: 1

Following that, force a recreation of your containers with docker-compose up --force-recreate --build and you should be good to go.

Gameplan

We plan for only a brief public beta so we can push new cloud images, documentation, and a setup script by the end of summer. We’ve been testing these changes internally for several months and the beta is quite stable. Thanks for your patience as this has been an off-kilter release cycle. We look forward to pushing a final release out ASAP.

10 Likes

Upgrade from V9 without any issue

  • I’ve recompiled to add LADP 2.9 dep
  • deployed on my kubernetes using helm

image

Congrats

3 Likes

Went from version 8 to 10 beta successfully. I have noted two things:

  1. I seem to have a growing queued jobs anomaly. I did a docker-compose exec redis redis-cli flushall around 4pm est and by 9 pm I had 233 queued jobs. No scheduled queries. Note I did truncate the database tables for queries, dashboards, events, widgets, query results, and changes.


2. Outdated Queries issue:

Does not seem to be impacting performance at this time.

2 Likes

Good news!

Are there any documentation changes for v10? Actually, most my queries to selfhosted Elasticsearch return

Error running query: argument of type ‘bool’ is not iterable

There will be, yes. But not pertaining to ES queries AFAIK.

I don’t think anything has changed with the ES connector. Which isn’t idea. The ES connector is in bad shape and we don’t recommend anyone use it for production work :frowning: Work was started on a new connector, but it’s blocked right now because we don’t use ES internally and can’t test it. If you can assist with the effort the PR is here: New Elasticsearch query runner by NicolasLM · Pull Request #4391 · getredash/redash · GitHub