v1.25.27 - LSM store compaction/flush synchronization Fix, Schema V2 redundant retries Fix, Backup Azure module configurable blocksize / concurrency improvement

antas-marcin released this 02 Dec 20:50

· 28808 commits to main since this release

e515414

Breaking Changes

none

none

chore: introduce github on-demand jobs by @antas-marcin in #6419
Add Weaviate vector DB header to JinaAI client by @antas-marcin in #6426
Fix lsmkv Compaction/Flush Synchronization Issues by @etiennedi in #6432
fix: flaky async indexing e2e tests by @antas-marcin in #6447
UX: Make GQL transform in GET/DELETE similar to CREATE by @kavirajk in #6475
backport(PR-6458): Fix redundant retries for non-multi-tenancy collections when querying tenants by @moogacs in #6489
Fix infinite _preload_progress log message when creating or loading an empty index by @etiennedi in #6492
Fix palm to google renaming with legacy vectors by @antas-marcin in #6499
Add support for X-Goog-* headers by @antas-marcin in #6500
Add environment overrides for azure blocksize and concurrency by @donomii in #6468
Add option of not waiting for self-deployed modules by @antas-marcin in #6501