Home Releases Core Releases

Core Releases

2025.03.18.3

  • Fixed bug affecting Always On duplication protection

2025.03.12.4

  • Fixed bug affecting Always On duplication protection

2025.03.10.1

  • Improved availability when updates are rolled out
  • Fixed bug affecting Always On duplication protection
  • Fixed issue where activation tests were contributing to Active Workflow count

2025.03.04.3

  • Added direct support for certificates to Node connection properties
  • Added caching runtime enabling it to continue operating during a storage outage
  • Hardened DropPoint bridge so that requests to DropPoints can continue to be serviced during a storage outage
  • Added automatic workflow activation of recent Workflows on all nodes in all clusters so that DR regions are kept hot
  • Fixed an issue in low-level maintenance process that controls log culling and new Node imports
  • Significantly improved quality of OpenAPI definition expport including ability to generate a definition for a subset of Workflows
  • Fixed an issue that affected accuracy of site-scoped metrics
  • Added CHIPS support to user cookie

2025.02.13.8

  • Added maintenance process to remove old installs

2025.02.13.5

  • Fixed bug where platform logs couldn't be downloaded
  • Fixed regression where legacy DropPoints were not able to connect due to signing replay attack mitigation
  • Fixed object ref exception that occurred on Node invokes when a platform shutdown was initiated

2025.02.11.1

  • Limit DropPoint package resends
  • New Tenant Features API
  • Implement versioning in the installer
  • Optimise indexing for activity log search

2025.02.05.1

  • Enhance audit trail to store action-specific properties
  • Added self-serve table store restore API (surfaced as a tool in Cluster screen)
  • Moved logging to dedicated store with fallback to a secondary store for higher resilience
  • Improved performance of indexed search

2025.01.21.3

  • Indexed search on Workflow logs
  • Added metrics showing ingest rates for log store
  • Included user ID in AI assistant calls
  • Randomize load balancing for similarly loaded hosts (previously the first host was always used)
  • Cleaned up shut down process to ensure dependencies know of their last chance to flush pending data
  • Add new API to query current user permissions
  • CLI function to save workflows to JSON file
  • Default branch name to env name on basic version control store

2024.11.25.2

  • Allow API-bound Workflows to auth via Bearer in addition to legacy Key=
  • Improve OpenAPI definition export
  • Internal monitoring enhancements to assist ops

v2024.11.12.4

  • Fixes to enhanced analytics
  • Fixes to CLI commands
  • Added Country field to User object
  • Added check for valid subscriptions before validating log retention

v2024.10.28.1

  • Added last sign in date to user metadata
  • Introduced enhanced analytics to assist with platform optimization
  • Fixes to CLI commands

v2024.10.17.4

  • Fix issue where prior Workflow version couldn't always be re-saved
  • Fix issue where cluster was not deemed primary when there were no other instances
  • Changed samples API to attempt dynamic samples first as we are moving away from static samples
  • Added help info call so tenant can see update ring and other help info
  • Fixed issue preventing @flowgear users from being added when subscription was at capacity
  • Multiple fixes & changes to CLI commands, primarily to support provisioning

v2024.09.16.2

  • Added rate limiter to platform logs
  • Added platform log throughput to metrics
  • Capture VM metrics by machine name not IP
  • Improve perf on realtime log load when busy
  • Shorten shutdown delay
  • Added min & max seen versions to /status call

v2024.09.10.7

  • Resolve 'path is not of a legal form' error that occurs in certain DropPoint invokes
  • Increase memory efficiency on messaging serializer
  • Partition DropPoints by cluster so that they only show online for clusters they are actually connected into
  • Resolve an issue that caused metrics to aggregate incorrectly
  • Added backing API's for My Activities
  • Fix issue where runtime could fail to shut down
  • Fix issue where error information contained unintended information when signing in
  • Added experimental reduce memory implementation (currently opt-in only)
  • Sensitized check to determine whether a cluster is in the primary region so Always On starts in secondary are delayed

v2024.08.22.5

  • Fix bug where invalid configurations were created
  • Hardening and additional parameters added to table store backup
  • Safety feature to prevent log culling when no explicit activity log age is defined

2024.08.20.8

  • Add support for Workflow invoke await timeout (assists in dealing with long running Workflows triggered from Run Now Pane)
  • Clean up critical logging to better differentiate errors that are outside of our control
  • Ensure potentially unsafe headers are redacted in FgRequestHeaders property

2024.08.08.3

  • Fixed an issue where the new serializer doesn't handle byte arrays identically with the old serializer

2024.08.06.5

  • Azure Table and Blob log truncates (report only mode released at this point)
  • Add log age limit to all product descriptors
  • Serve platform log downloads as file rather than open in browser page
  • Disallow delete Workflows that have Always On enabled for one or more Environments
  • Fix bug where imported Node would be marked latest even if there was a higher version
  • Added a delay before starting maintenance processes like log truncation, also acquiring a tenant-wide lock before starting
  • Fixed obj ref on shut down
  • Initial implementation of memory reduce feature, requires explicit opt-in
  • Added backend to support Key/Value management features
  • Added legacy compatibility for basic auth on API invokes
  • Cleaned up MessageHub state detection and socket close handling
  • Implemented better close handling on DropPoint transport
  • Fixed race condition affecting DropPoint ping timer
  • Aligned parallel sub-Workflow activation mode with legacy platform (ensures that Nodes execute from RunNow output)
  • Added pseudo-HA support for DropPoint (allows two instances to connect with the same identity, randomly routes requests to them)
  • Enhanced runtime manager shut down so that it catches tasks that have just started up (this issue was preventing runtime shutdowns in some cases)
  • Added platform log for when a parallel sub-Workflow invoke is delayed due to insufficient Active Workflow capacity
  • Resolved metrics issue where data generated by earlier runtime instance wasn't shown
  • Added support for secondary cluster and ensure that Always On workflows fire up in secondary when it its VM's are alloc'd to the lowest /24 of known VM's across all clusters
  • Added query option to /status API to ease assessment of whether failover has taken place
  • Replaced MessageHub destructive serializer with a non-destructive but still memory-efficient serializer
  • Fixed an issue where an overflowing error message on a Node prevented that log batch from being persisted
  • Ensures that common errors like thread aborts that are unobserved aren't treated as critical
  • Fixed issue where route segments that contain values with escaped chars like %20 weren't handled correctly

2024.07.02.13

  • Fix object store regression issues (overwrite of existing objects)
  • Fixed infinite loop issue due to race condition
  • Fixed unobserved exception
  • Fixed issue where DropPoints were reported online but were not
  • Added workflow recycle (unload/reload for Workflows executing at a Trigger)
  • Increased runtime responsive timeout
  • Added compatibility option to ensure dates serialize identically with old platform (affects Script Nodes and data mapping expressions that pluck date/time scalars)
  • Added Node auto update with delay