- Fixed an issue when new columns sometimes not added to target table
- Fixed an issue when the query runner returned incorrect value when multiple queries running in one transaction
- Switch jsonschema to use Draft7Validator
- Fix loading tables with space in the name
- Generate compressed CSV files by default. Optionally can be disabled by the
no_compression
config option
- Log inserts, updates and csv size_bytes in a more consumable format
- Use SHOW SCHEMAS|TABLES|COLUMNS instead of INFORMATION_SCHEMA
- Support usage of reserved words as table names.
- Support custom logging configuration by setting
LOGGING_CONF_FILE
env variable to the absolute path of a .conf file
- Change default /tmp folder for encrypting files
- Make AWS key optional and obtain it secondarily from env vars
- Add temp_dir optional parameter to config
- Fixed issue when JSON value not sent correctly
- Load binary data into Snowflake BINARY data type column
- Add missing module
python-dateutil
- Review dates & timestamps and fix them before insert/update
- Pinned stable version of
urllib3
- Pinned stable version of
botocore
andboto3
- Fixed issue when extracting bookmarks from the state messages sometimes failed
- Bump
snowflake-connector-python
to 2.0.3
- Fixed an issue when number of rows in buckets were not calculated correctly and caused flushing of data at the wrong time with degraded performance
- Fixed an issue when sometimes the last bucket of data was not flushed correctly
- Bump
snowflake-connector-python
to 2.0.1 - Always use secure connection to Snowflake and force auto commit
- Add
flush_all_streams
option - Add
parallelism
option - Add
max_parallelism
option
- Emit new state message as soon as data flushed to Snowflake
- Log SQLs only in debug mode
- Further improvements in
information_schema.tables
caching
- Improved and optimised
information_schema.tables
caching
- Caching
information_schema.tables
to avoid long running SQLs in snowflake - Instead of DROPPING exiting column RENAME it
- Add
data_flattening_max_level
option
- Optimised queries to
information_schema.tables
- Create
_sdc_deleted_at
asVARCHAR
to avoid issues caused by invalid formatted date-times received from taps
- Manage only three metadata columns:
_sdc_extracted_at
,_sdc_batched_at
and_sdc_deleted_at
- Initial release