11.34.121 Platform Release
Released November 16, 2026
This release is a patch update for the 11.34 platform release.
Addressed issues
-
BQL: We've improved performance of Finding searches that use CONTAINS ANY and EXISTS on Asset, which previously could time out after 5 minutes.
-
Compliance: We’ve improved compliance status accuracy so expired exception requests no longer incorrectly mark items as Out of SLA.
-
Compute: We've improved compute performance by reducing timer contention when reading records.
-
Compute: We've improved relationship persistence performance during compute by reading computed records in parallel.
-
Consolidation: We’ve enhanced Max/Min consolidation criteria so attributes with mixed-type source values are handled gracefully and skipped, keeping consolidation running smoothly.
-
Consolidation: We've improved consolidation by making the storage open-file and background-job limits configurable.
-
Consolidation: We've improved consolidation flexibility by making source-record and consolidated-object reader parallelism configurable in generated consolidation flows, where it was previously fixed.
-
Consolidation: We've improved consolidation performance by running the identity-ordering phase in parallel.
-
Consolidation: We've improved consolidation performance by removing a bottleneck that forced attribute mapping to run serially.
-
Consolidation: We've improved consolidation performance by loading consolidated objects in parallel instead of through a single thread.
-
Consolidation: We've improved consolidation performance by removing lock contention during the convert phase, so it scales with the number of writers.
-
Consolidation: We've improved consolidation performance by running the early consolidation phases in parallel instead of on a single thread.
-
Consolidation: We’ve improved batch delete reliability by retaining task results until they’re read, so consolidation stays on the fast path instead of falling back to one-at-a-time deletes.
-
Consolidation: We've improved consolidation memory usage by keeping only record IDs in memory, so memory use no longer grows with the full contents of every source record on very large datasets.
-
Data integration: We’ve enhanced data integration so connectors that return a list of date-time values sync smoothly.
-
Data Warehouse: We've improved data warehouse snapshot performance by defaulting parallelism to the number of available CPU cores.
-
Data Warehouse: We've improved data warehouse snapshot performance by exporting relationships more efficiently, significantly reducing snapshot time on very large datasets.
-
Lifecycle: We’ve improved lifecycle stamp and purge on large datasets by streaming record IDs instead of holding them in memory, keeping memory use low and steady.
-
Manual entry: We’ve enhanced manual creation of consolidated records (for example, an Application) so it completes smoothly.
-
Performance: We've improved matching performance by avoiding evaluating every rule for each record, which was costly for configurations with a very large number of rules.
-
Performance: We've improved consolidation write performance by using lighter-weight, equality-only indexes for long text attributes, reducing index overhead and CPU use.
-
Performance: We've improved write performance on large datasets by reducing per-write index checks, lowering CPU use.
-
Platform: We've improved performance on large datasets by addressing a low default limit on open storage files, and unsafe file-limit warnings are now logged during schema migration.
-
Platform: We’ve improved startup recovery after an unexpected server crash so schema indexes stay in place, making recovery faster.
-
Tickets: We’ve enhanced automated creation and update of vulnerability tickets during orchestration with batched ticket relationship writes, for smoother, more reliable runs.
-
Tickets: We’ve improved ServiceNow problem task creation so a timeout no longer leads to duplicate tasks on retry.
-
Tickets: We’ve enhanced ad hoc ticket creation for users whose read access depends on cluster membership, so it completes quickly even when the cluster contains a large number of records.