Jitsu 1.40 is here! Better User Recognition, Lower Memory usage and more

Ildar Nurislamov

VP of Engineering
February 22nd, 2022

We're pleased to announce Jitsu v1.40 with significant improvements for Retroactive User Recognition feature, server memory consumption, and new Slack notifications for synchronization tasks statuses.

User Recognition overhaul

We have completely overhauled the Retroactive User Recognition feature. Now it applies to API token level. Recognized events repeat exactly the same pipeline as original events, as a result: User Recognition now correctly works with destinations in batch mode. User Recognition now correctly works with data mappings and transformations.

Starting from 1.40, User Recognition ID nodes customization are no longer supported on destination level. Please use global level settings

Slack Notification

Users can now add Slack notifications for synchronization task status per project or globally for the Jitsu server.

Memory Consumption improvements

  • User Recognition no longer causes indefinite growth of inmemory queue on setups with high events count.
  • V8 engine heap size limited to 100mb preventing constant memory usage growth on setups using javascript transformation
  • Redis source now loads data in chunks - reducing overall memory consumption and preventing OOM crashes on huge volumes.

New version of @jitsu/sdk-js

We released a new version of @jitsu/sdk-js that allows to add custom headers to tracking requests. Read more about JS SDK configuration here (look for custom_headers parameter!)

Other Changes

  • Clickhouse driver updated to the latest version
  • Facebook API version upgraded to v13.0
  • Improved responsiveness of Configurator UI by migration to objects v2 api
  • Bugfixes

About Jitsu

Jitsu is an open-source data integration platform offering features like pulling data from APIs, streaming event-base data to DBs, multiplexing and many others.
© Jitsu Labs, Inc

2261 Market Street #4109
San Francisco, CA 94114