Instruction

Preparation

To update, check that the version is set in the .env configuration file:

~# grep VERSION .env
SW_FRONTEND_VERSION=v5.9.2
SW_BACKEND_VERSION=v5.9.2
SW_HELPCENTER_VERSION=v5.9.2

If you plan to use two-factor authentication, you must add a new feature flag 2fa.

Also, at the same step, you can enable optimization of the knowledge base portal for LLM.

Update

Perform the update with a system restart:

docker compose pull
docker compose down
docker compose up -d

After the update

After all containers start and the service becomes available through the web interface, complete the post-upgrade steps:

1. Run the post-start command to update static files, search indexes, and perform other necessary procedures:

docker exec --env-file .env -it swarmica-django-1 python manage.py post_start

This command may take a long time, depending on the amount of data in the system.

1. Run the command to rename the chat_status field to status:

docker exec --env-file .env -it swarmica-django-1 python manage.py migrate_chat_status_to_status

1. Run the command to migrate the auto-assignment status to the status field:

docker exec --env-file .env -it swarmica-django-1 python manage.py migrate_autoassign_to_status

1. If you use YandexGPT, then in this release you need to reconfigure it:

4.1. Switch the model from YandexGPT-Lite to YandexGPT:

docker exec --env-file .env -it swarmica-django-1 python manage.py migrate_yandex_gpt_model

4.2. Reset the YandexGPT state to "clean":

docker exec --env-file .env -it swarmica-django-1 python manage.py wipe_yandex_cloud

After all post-upgrade steps are completed, you will need to go to the graphical interface > Settings > Integrations > AI Providers and re-enable AI assistants.

1. Run the calculation of the SLA non-compliance coefficient, which is used to sort tickets during auto-assignment:

docker exec --env-file .env -it swarmica-django-1 python manage.py recalculate_sla

1. Run the cleanup of unused data:

docker system prune -a

✨ Key improvements

🔐 Security and access

• The ability to use 2FA via TOTP and Email
• Blacklist of compromised JWT tokens
• Masking of passwords, tokens, and other sensitive data in the API
• Secure form for entering passwords, tokens, and other sensitive data

🧠 Working with articles

• The ability to translate the entire article (AI)
• The ability to create 30+ tabs in an article
• Auto-anchors for headings
• Events for changing the statuses and segments of articles
• Advanced SEO settings for articles
• Optimization of the knowledge base for LLM

📬 Automation and templates

• Support for Jinja in email automations (to / cc / bcc)
• Support for Jinja variables in tickets

📊 Reports and filters

• Added new objects and measurable entities
• Support for isnull for products/layouts/versions/platforms
• Modifiers for Hour and Day of the week for Date and Time fields in reports + the ability to set the time zone for these modifiers

💬 Messengers and channels

• Deepened integration with MAX
• Added settings, links, and support in the widget

🔧 Bug fixes

🎨 Fixed issues with the editor: video, text formatting, undo, drop-down menus, and styles.
📊 Report display: locales, dates, SLA counters, and data display.
🧾 Fixed bugs in auto-assignment logic and event processing.
💬 Improved stability of messengers and comment processing.
⚙️ Improved system stability and performance: cache, API, migrations, and background processes.