Releases · man-group/ArcticDB

Update logo (#2154)
Implement batch_update (#2100)
Only publish benchmark results for master builds (#2157)
Fix issue with running only selected tests (#2155)
Read and write snapshot metadata from the correct place (#2161)
Multiple fixes for STS token refresh (#2115)
Extend update docs to mention possible data fragmentation (#2164)
Update Arrow so that ArcticDB can be compiled with Clang 19 (#2134)
Read arrow data (#2141)
Small improvements to test_compatibility.py (#2163)
DBA tools for ReliableStorageLock (#2152)
8234437666 Bound the number of data segments kept live in memory during compaction and sort_merge (#2167)
8066268338 Key segment pair refactor - do not move the segment out of the KeySegmentPair as it is a shared resource (#2166)
Fix debug symbols not attaching to release and release script failing (#2168)
Remove unnecessary clone following the key segment pair refactor (#2169)
Enhance MockS3Client to support real client delegation and configurable failures (#2158)
Simplify our unpickling logic and add tests for Python 2 pickles (#2160)
Enhancement: Implement for NFS and Azure storages (#2172)
Automated Release scripts (#2174)
Include wheels in releases (#2180)

The wheels are on PyPI. Below are for debugging:

Assets 30

07 Feb 14:41

github-actions

v5.2.3

39460a6

v5.2.3

🐛 Fixes

Fix to bound the number of data segments kept live in memory during compaction and sort_merge by @poodlewars in #2170
Fix debug symbols not attaching to release and release script failing… by @ognyanstoimenov in #2171

Full Changelog: v5.2.2...v5.2.3

The wheels are on PyPI. Below are for debugging:

Contributors

ognyanstoimenov and poodlewars

Assets 16

03 Feb 14:10

github-actions

v5.2.2

0e31391

v5.2.2

Fixes

Read and write snapshot metadata from the correct place PR #2161 .

Snapshot metadata is an optional structure of extra information saved with a snapshot, created when you call,

from arcticdb import Library
lib: Library

lib.snapshot("snap", metadata=["example", "metadata"])

lib.list_snapshots()
# {"snap": ["example", "metadata"]}

If you never call snapshot with metadata= then you are not affected by this issue. Metadata associated with a symbol is not affected by this issue.

Between v4.5.0 and v5.2.1 we mistakenly changed to save snapshot metadata in a new data format. This affects the snapshot metadata in the dict values returned by list_snapshots. Those versions incorrectly return None for snapshot metadata written by earlier versions, even if that metadata does exist.

This fix has us now check both data formats for snapshot metadata, so ArcticDB v5.2.2 onwards is compatible with snapshot metadata written by any version. We also revert to writing the metadata in its old location from v5.2.2 onwards. This means that ArcticDB clients between v4.5.0 and v5.2.1 will need to upgrade to be able to read snapshot metadata written by clients outside of those versions (otherwise they will see the metadata as None). ArcticDB clients older than v4.5.0 need to upgrade to v5.2.2+ to be able to read snapshot metadata written by versions between v4.5.0 and v5.2.1.

Full Changelog: v5.2.1...v5.2.2

Assets 2

29 Jan 16:46

poodlewars

v5.2.1

c60bf1f

v5.2.1

🐛 Fixes

Use Pandas unpickling to handle Pandas 1 vs Pandas 2 API differences better . This only affects reading Pandas structures saved as metadata with an ArcticDB symbol. (#2151)
Don't warn about missing keys when reading symbol ref as it is common to try to read a non-existent symbol (#2153).
- This fixes a minor regression in v5.2.0 that caused noisy logging output like Failed to find segment for key 'r:aaa' : No response body..

Full Changelog: v5.2.0...v5.2.1

The wheels are on PyPI.

Assets 2

27 Jan 14:54

github-actions

v5.2.0

18bb7bb

v5.2.0

🚀 Features

Introduce the block version ref key (#1969)
Add AWS STS authentication support (#1884)
Python 3.12 and Python 3.13 support (#1945) (#2016)
Numpy 2 support (#2050)
Reliable storage lock (#2014)
Storage mover (#2039)
Add S3 STS proxy support (#2072)
Implement origin for pandas resampling (#1962)
Refactor to storages to support async reads (#2012)
Chunk up incomplete segments by rows when they are staged (#2117)
read_batch performance improvements - now up to 10 times faster

🐛 Fixes

Fix decoding of fields with >2^16 blocks (#2089)
Handle very old normalization metadata RangeIndexes (#2118)
Finalize staged data memory use improvements (#2013)
Fix handling of empty DF in pandas 1.0 (#2010)
Fix string reference count leak (#1998)
Fix version release not attaching symbols for debug (#2018)
Delete staged segments after writing vref key (#2037)
Improve the performance of update by parallelising reads. Implement internal async update method. (#2087)
Performance regression when requesting a timestamp before the earliest version (#2076)
Using compact incomplete on a library with dynamic schema with a named index can result in an unreadable index (#2116)
Notimplemented handling (#2108)
Fix race between list_versions and delete_snapshot on NFS (#2092)

Full Changelog: v5.1.3...v5.2.0

The wheels are on PyPI. Below are for debugging:

Assets 13

21 Jan 11:36

github-actions

v5.1.3

f4108f1

v5.1.3

What's Changed

Bugfix 4897570890562900007: Fix decoding of fields with >2^16 blocks by @alexowens90 in #2088
Handle very old normalization metadata RangeIndexes (#2118) by @alexowens90 in #2120
5.1.x Add exception when staging with different namd indexes by @G-D-Petrov in #2126
Update the version for upload and download artifact GH actions to v4 by @vasil-pashov in #2131
Storage mover port (#2039) by @vasil-pashov in #2127