Change Record
The current version of DataFlux Func is 6.x, and all previous versions have been archived. Only downloads are provided; no further updates or support will be given.
- View DataFlux Func 6.x change log
- View DataFlux Func 5.x change log
- View DataFlux Func 3.x change log
- View DataFlux Func 2.x change log
- View DataFlux Func 1.x change log
- Download old versions
For the version correspondence relationship between "Independent Deployment Edition DataFlux Func" and "Guance Attached Func", please refer to:
1.7.31 (2022-11-02)
- The Kafka default Group ID has been changed to a fixed name (previously it was random).
1.7.30 (2022-10-28)
- Code adjustments.
1.7.29 (2022-10-28)
- Automatic refresh of subscription information in the connector subscription interface.
- Fixed known bugs.
1.7.27 ~ 1.7.28 (2022-10-28)
- UI optimization.
1.7.26 (2022-10-27)
- Adjusted Helm upgrade version issues.
- Added recent consumption time records to the connector subscription interface.
1.7.25 (2022-10-26)
- Fixed PVC issues during Helm deployment.
- Optimized subscribers.
1.7.24 (2022-10-25)
- Adjusted/Optimized code.
1.7.23 (2022-10-24)
- Adjusted some default configurations.
1.7.22 (2022-10-24)
- Adjusted docker-stack.yaml, added default replicas configuration items.
1.7.21 (2022-10-24)
- Optimized sidebar tooltips to prevent overly long function definitions from causing content width issues.
1.7.20 (2022-10-23)
- Optimized installation interface UI.
- Optimized file manager UI.
1.7.19 (2022-10-19)
DFF.CONN(...)
method for obtaining DataKit/DataWay connector objects addedtimeout
parameter.- Adjusted PYPI image source.
- Adjusted subscription processing, each topic creates a separate connection.
- Refactored function call parameter construction handling.
- Optimized UI.
1.7.18 (2022-10-14)
- Fixed known bugs.
1.7.17 (2022-10-14)
- Added authentication support to Kafka connectors.
1.7.16 (2022-10-12)
- Optimized Kafka subscription processing.
1.7.15 (2022-10-12)
- Optimized Kafka subscription processing.
- Optimized performance issues with some interfaces.
- Adjusted task information cleanup processing.
1.7.14 (2022-10-11)
- Fixed an issue where connectors would report errors after stopping subscriptions.
1.7.13 (2022-10-09)
- Optimized administrator tool tips.
- Optimized code marker jump handling.
- Fixed known bugs.
1.7.12 (2022-09-30)
- Optimized search box logic in the script market.
1.7.11 (2022-09-29)
- UI adjustment.
1.7.10 (2022-09-27)
- Added custom website Favicon.
1.7.9 (2022-09-22)
- 【New Feature】 Added Kafka connector support.
- 【New Feature】 Increased system configuration options:
- Support for custom website titles.
- Support for custom website logos.
- Support for custom documentation addresses.
- Added
STAR
code comments. - Optimized dropdown search boxes using similarity algorithms for filtering/sorting.
- Optimized connector subscription processing.
- System configuration supports adjusting UV thread pool parameters.
- Other code/UI optimizations.
- Fixed known bugs.
1.7.8 (2022-09-15)
- Fixed the issue where specifying
search_after=[]
inDataKit.query(...)
did not take effect.
1.7.7 (2022-09-11)
- Supported creating DataWay operation objects directly via
DFF.DATAWAY(...)
. - Added Python/Node version number display on the about page.
1.7.6 (2022-09-10)
- Code adjustment.
1.7.5 (2022-09-09)
- Compatible with some versions of DataKit that did not correctly mark the response body's Content-Type.
1.7.4 (2022-09-08)
- Added
DFF.SIGN
method.
1.7.3 (2022-09-08)
- Fixed the potential UI freeze issue under certain conditions.
- Adjusted fonts.
1.7.2 (2022-09-08)
- Code adjustment.
1.7.1 (2022-09-08)
- Fixed some UI errors.
1.7.0 (2022-09-07)
- 【New Feature】 Added system configuration page (configurable top notification bar, monitoring data reporting to Guance).
- 【New Feature】 Added prohibition of script import/installation configuration item
_DISABLE_SCRIPT_SET_IMPORT
. - Added editor sidebar function association authorization links, automatic trigger configurations, quick batch processing view pop-up windows.
- Added function parameter value copy functionality.
- Added English README.
- Renamed “data source” to “connector”.
- Terminal output logs are uncolored by default.
- Fixed script editor sidebar jump issues.
- Optimized various dropdowns/selectors.
- Removed historical legacy code.
- Updated third-party package versions.
- Other code/UI adjustments.
- Other UI adjustments.
1.6.14 (2022-07-28)
- Optimized editor, supporting theme/font size configuration directly on the editor page.
- Optimized function/authorization link API documentation page display.
- Optimized script market error prompts.
- Optimized processing speed of
GET /api/v1/metrics
interface. - Optimized other interface displays.
1.6.13 (2022-07-26)
- 【New Feature】 Added jump-to-class, TODO tag quick selector in the editor. Supported TODO tags include:
XXX
TEST
TODO
BUG
FIXME
HACK
1.6.12 (2022-07-19)
- 【Important】 Starting from this version, the license agreement has been changed to AGPLv3.
- Optimized function execution output, adding peak memory allocation information.
- Editor parameters support multi-line editing.
- Corrected introduction page text.
- Optimized editor sidebar behavior.
- UI optimization.
1.6.11 (2022-06-28)
- Updated CodeMirror version, fixed code scrolling lag issues.
- Adjusted some pages.
1.6.10 (2022-06-21)
- Fixed file upload-related processing failure issues (this problem was introduced in version 1.6.9).
1.6.9 (2022-06-17)
- When authorized links submit non-JSON/text data types, Base64-encoded binary data is used instead.
1.6.8 (2022-06-17)
Optimized DataKit.query(...)
method.
1.6.7 (2022-06-17)
- Fixed the issue with incorrect curl code generation in API examples.
1.6.6 (2022-06-17)
- Authorized link calls now support non-JSON data types.
1.6.5 (2022-06-15)
- Fixed Helm deployment issues.
- Partial code adjustments.
- Fixed the issue where logs were not correctly colored on the log page.
1.6.4 (2022-06-14)
- Fixed the issue where DDL was not the latest version during a fresh installation.
1.6.3 (2022-06-14)
- Fixed Helm deployment issues.
- Optimized configuration file compatibility.
- Optimized the rotation deletion handling logic for log tables.
1.6.2 (2022-06-08)
- Fixed the issue where queues were blocked due to unresponsive data sources when adding/modifying data sources.
1.6.1 (2022-06-07)
- Adjusted the
docker-stack.yaml
file, added container log volume configuration restrictions.
Upgrading from older versions
Since the DataFlux Func upgrade program does not modify the already installed docker-stack.yaml
file,
users upgrading from older versions to this version need to manually add relevant log configurations in {installation directory}/docker-stack.yaml
and restart the system.
Reference configuration as follows:
YAML | |
---|---|
1 2 3 4 5 6 7 8 9 10 11 12 13 |
|
Before modifying docker-stack.yaml, please back up this file
1.6.0 (2022-05-23)
- Optimized script market UI.
- Optimized
DFF.STORE
andDFF.CACHE
handling. - Optimized performance of log table rotation deletion.
- Optimized caching and package import strategies.
- Adjusted script log output format (removed year, added milliseconds).
- Optimized editor sidebar highlight jump logic.
- Authorized links support adjusting task record quantity.
- Fixed other known issues.
1.5.17 ~ 1.5.18 (2022-05-07)
- Added Helm-related configurations.
1.5.16 (2022-05-05)
- Added
DFF.CACHE.keys(...)
to retrieve cache key lists. - Adjusted some UI styles.
- Fixed known issues.
1.5.14 ~ 1.5.15 (2022-05-01)
- Added markdown library.
- Fixed inconsistencies between API interfaces and configuration files regarding
task record quantity limit
.
1.5.13 (2022-04-26)
- Fixed database backup files having inconsistent times.
- Refactored automatic database backup handling.
- Adjusted installation interface.
- Admin user passwordHash defaults to empty and supports login using username as password when passwordHash is NULL.
- Added open-source component description document (OPENSOURCE.md).
1.5.12 (2022-04-23)
- Authorized links added recent task records.
- Optimized automatic trigger configuration/batch processing list display.
- Recent task records optimized.
1.5.11 (2022-04-22)
- Search box allows re-execution of searches without changing search content by clicking the icon.
- Fixed the issue of history record rollback errors in the search box.
- Different log volumes displayed in different colors.
- Task records added log line count display.
- Adjusted dependency packages.
- Fixed known issues.
- Adjusted some UI styles.
1.5.10 (2022-04-12)
- In PIP tools, use University of Science and Technology of China as the default mirror.
1.5.9 (2022-04-11)
- Fixed installation/importing script set issues.
1.5.8 (2022-04-11)
- 【New Feature】 System administrators can forcibly edit locked scripts/script sets.
- Locked scripts/script sets provide lock owner information.
- Removed the status bar on the code editor page.
1.5.7 (2022-04-10)
- Fixed the issue where task records were not retained according to the configuration.
- Optimized some UI elements.
1.5.5 ~ 1.5.6 (2022-03-30)
docker-stack.yaml
configuration file added--tls-version=TLSv1.2
option formysql
service.- Fixed the issue where functions returning empty lists/dicts resulted in actual authorized links returning empty content.
- Fixed the issue where script IDs were too long when creating scripts.
- Adjusted non-SA user page permission controls.
- Optimized some UI elements.
Regarding the resolution of the ERROR 2026 (HY000)
error when publishing/executing scripts/functions
Since Ubuntu 20.04 requires SSL version v1.2 at minimum, after upgrading an old version of DataFlux Func to this version, you may encounter the following exception when publishing/running scripts:
Text Only | |
---|---|
1 |
|
DataFlux Func version 1.5.5 and later includes built-in automatic configuration modification handling, so normal upgrades should suffice.
If the issue persists, you can attempt the following manual fix:
- Open the
<installation directory>/docker-stack.yaml
file (default location is/usr/local/dataflux-func/docker-stack.yaml
) - Add the
--tls-version=TLSv1.2
parameter to theservice
/mysql
/command
section of the yaml file.
Reference modifications as follows:
Text Only | |
---|---|
1 2 3 4 5 |
|
1.5.4 (2022-01-20)
- Optimized code to improve opening speed of automatic trigger configuration list/batch processing list pages.
1.5.3 (2022-01-19)
- Reduced maximum task record count from 10000 to 500 (too many task records may cause performance issues).
1.5.2 (2022-01-19)
- 【New Feature】 Added
HOST_HOSTNAME
environment variable inside containers for retrieving the hostname of the host machine in scripts. - Fixed known integrated Crontab execution issues.
About the HOST_HOSTNAME
Environment Variable
In scripts, you can use the following code to retrieve the hostname of the host machine:
Python | |
---|---|
1 2 |
|
This feature is implemented by configuring environment variables in the Docker Stack YAML file. If upgrading from an old version of DataFlux Func, you need to manually modify the YAML file to add this environment variable.
Older versions of DataFlux Func can also achieve the above functionality by directly modifying the Docker Stack YAML file. Modify as follows:
- Open the Docker Stack configuration file for DataFlux Func
Bash | |
---|---|
1 |
|
- Add the following
environment
configuration to all Workers (reference code):
YAML | |
---|---|
1 |
|
- Restart DataFlux Func
1.5.1 (2022-01-18)
- Optimized some UI interfaces.
- Fixed the issue where batch processing could not be called using the simplified POST form.
1.5.0 (2022-01-18)
If you use the 'batch processing' module in DataFlux Func, please upgrade to this version as soon as possible.
- 【New Feature】 Automatic trigger tasks and batch processing tasks support specifying the number of task records (default 10, selectable 1~10000).
- 【New Feature】
DFF.API(...)
addsfixed_task_info_limit
parameter (default 10, selectable 1~10000). Functions configured with this parameter retain a fixed number of task records when executed as automatic triggers or batch processes (similar tofixed_crontab
).
Number of Subtask Records
When the executed function contains a large number of subtasks, the system ensures that the task records of the last main function and all its subtasks will be recorded.
Assume: Automatic trigger configuration task record count is set to 3, and currently there are 3 task records. The executed function calls 10 additional sub-functions through
DFF.FUNC(...)
.Then: The final task record count is
3 + 10 = 13
task records.
Users do not need to worry about insufficient subtask records due to setting too few task records (limited to the last main task only).
1.4.15 (2022-01-05)
- UI Adjustment: For performance reasons, the "Function Cache Manager" and "Function Storage Manager" interfaces only support previewing data less than 20KB.
- Fixed known UI issues.
1.4.13 ~ 1.4.14 (2021-12-30)
- UI Adjustment: Directly display task record Exception information in the list.
- Script Execution: Adjusted import behavior of user scripts to conform to native Python import statement handling.
1.4.12 (2021-12-21)
- Fixed known UI issues.
1.4.11 (2021-12-20)
- Fixed the issue of some MySQL versions unable to connect.
1.4.10 (2021-12-15)
- Fixed unreasonable chart display issues in system metrics.
1.4.9 (2021-12-14)
- Fixed the issue of emoji not being correctly output in the code editor output box.
1.4.8 (2021-12-13)
- Adjusted automatic trigger configuration/batch processing task records:
- Each log line is limited to 5,000 characters (longer lines will be truncated).
- Each task log is limited to 50,000 characters (exceeding this will retain the first 10,000 and the last 40,000 characters).
- Automatically triggered tasks/batch processes created via API interfaces (such as Guance monitoring processing) retain up to the most recent 10 task records.
- Automatically triggered tasks/batch processes manually created via the DataFlux Func interface retain up to the most recent 300 task records.
- Automatic trigger configuration/batch processing task records support clearing.
Reference: The full Chinese edition of "The Old Man and the Sea" is approximately 40,000 words.
1.4.5 ~ 1.4.7 (2021-12-09)
- Fixed the issue of logs not being displayed in automatic trigger configuration/batch processing task records.
- Optimized script market display.
- Optimized UI.
1.4.2 ~ 1.4.4 (2021-12-02)
- Optimized UI.
1.4.1 (2021-12-01)
- Improved support for viewing associated tasks in automatic trigger tasks and batch processing tasks.
1.4.0 (2021-12-01)
- 【New Feature】 Added support for viewing associated tasks in automatic trigger tasks and batch processing tasks.
- Optimized UI.
1.3.14 (2021-11-25)
- Added arm64 Oracle support.
1.3.13 (2021-11-25)
- SQL-type data sources no longer provide database-level browsing, instead displaying database tables directly.
1.3.12 (2021-11-25)
- 【New Feature】 Added abnormal request recording functionality.
- No longer uses custom Chinese fonts.
- Improved some Chinese and English translations.
- Fixed the issue where the GET /api/v1/func-list interface failed to correctly assemble function call URLs.
- Fixed page memory leakage problems.
- Fixed other known issues.
1.3.11 (2021-11-11)
- Optimized file manager, supporting direct installation of Wheel packages and selection of PIP mirrors.
1.3.10 (2021-11-11)
- 【New Feature】 File manager supports direct installation of
.whl
packages. - 【New Feature】 Added
GET /api/v1/metrics
interface to output metric data compliant with the OpenMetric protocol. - Fixed performance issues caused by returning large amounts of data.
- Fixed inability to install third-party packages with underscores in names in the PIP tool.
- Default sa user name changed to
Administrator
. - Timeout for directly running functions in the editor reduced from 60 seconds to 45 seconds, ensuring it is shorter than the default nginx timeout.
- Function run record table added automatic deletion of outdated content handling to reduce database disk usage.
- Adjusted some UI.
- Fixed other known issues.
- Prompt Worker basic processing performance.
1.3.9 (2021-10-27)
- Fixed issues saving/testing connectivity for DataKit data sources.
- Some UI adjustments.
1.3.7 ~ 1.3.8 (2021-10-20)
- Optimized display effects for some UI elements at 1280px width.
1.3.5 ~ 1.3.6 (2021-10-19)
- 【New Feature】 Added "DataFlux Func Sidecar" functionality.
1.3.4 (2021-10-15)
- Fixed the issue where DFF.FUNC(...) processing failed to record logs.
- Some code adjustments.
- Some UI adjustments.
1.3.3 (2021-10-14)
- API Authentication: Header field authentication does not distinguish case for field names.
- Adjusted UI display, minimum page width is 1280; increased authorized link, automatic trigger configuration, and batch processing function information display width.
- Fixed the issue where the simple debugging panel failed to correctly display error messages.
- Added example code hints to the authentication function configuration in the API authentication configuration.
1.3.2 (2021-10-07)
- 【New Feature】 Added "API Authentication" module functionality, supporting:
- Verifying field values in the request Header, Query, or Body.
- HTTP Basic Authentication.
- HTTP Digest Authentication.
- Custom function authentication.
- Fixed the issue where the UI sidebar had an excessively large click range, making it easy to accidentally click adjacent items.
- Adjusted some wording.
- Optimized internal JSON handling.
- Added test cases, fixed known issues.
1.3.1 (2021-09-22)
- Fixed the issue of the first-time installation interface freezing.
1.3.0 (2021-09-17)
- 【New Feature】 Added built-in DFF.RESP_LARGE_DATA(...) functionality to improve performance when returning large data to clients.
- Fixed the issue where file uploads generated
undefined
folders. - File uploads renamed automatically appended with -2, -3 suffixes.
1.2.1 (2021-09-16)
- Adjustments made to copywriting, domain names, etc., to align with the renaming of the "DataFlux" product to "Guance" (supplementary update).
1.2.0 (2021-09-16)
- Adjustments made to copywriting, domain names, etc., to align with the renaming of the "DataFlux" product to "Guance" (DataFlux Func itself does not participate in the renaming).
1.1.17 (2021-09-15)
- Replaced the Guance edition logo.
1.1.16 (2021-09-15)
- Fixed the issue of incorrect display of function storage content.
- To avoid front-end performance issues, the function cache/storage manager no longer performs code highlighting.
- Optimized long text viewing processing logic.
1.1.15 (2021-09-15)
- Optimized the UI of the function cache manager.
1.1.14 (2021-09-15)
- Fixed the issue where the function cache manager could not function properly when colons were included in the function cache key names.
1.1.13 (2021-09-14)
InfluxDBHelper
openedwrite_point(...)
andwrite_points(...)
methods for writing data points.
1.1.12 (2021-09-11)
- Fixed the issue where batch processing task information was not saved correctly.
- Fixed the issue where "disabled" was written as "hidden" in the automatic trigger configuration list.
- Celery Beat internal output log level changed to
error
.
1.1.11 (2021-09-09)
- Added default value parameter
default
to theDFF.ENV(...)
function. - Added default value parameter
default
to theDFF.CONFIG(...)
function.
1.1.10 (2021-09-08)
- Fixed known issues.
1.1.9 (2021-09-07)
- Added
REDIS_USE_TLS
option (true
/false
) to the configuration file to meet AWS ElasitCache's requirement to use TLS when configuring passwords.
1.1.8 (2021-09-06)
- Fixed known issues.
1.1.7 (2021-09-04)
- Fixed known issues.
1.1.6 (2021-09-03)
- Optimized database JSON processing compatibility with Mariadb.
- Logs added
hostname
field output; addedLOG_TIMEZONE
configuration to specify the timezone for log times. - Optimized custom ID configuration display, added checks for IDs to only contain alphanumeric characters, underscores, and dots.
- Optimized network error prompts to prevent continuous network error prompts.
- Optimized monkey patching prompts.
- Clearing caches clears prompt statuses.
1.1.5 (2021-08-25)
- Fixed some copywriting errors.
1.1.4 (2021-08-24)
- Installation script added automatic check for the previous installation directory.
1.1.3 (2021-08-24)
- Fixed the issue of pages being identified as English by Chrome browser.
1.1.2 (2021-08-24)
- Added
aarch64
architecture support.
1.1.0rc1 ~ 1.1.0rc72 (2021-08-19)
- Released stable pre-release version.
- Imported/exported authorization links, automatic trigger configurations, and batch processes only replace data with the same ID.
- API call examples no longer provide "flat" form parameter passing.
- Added page-level code editor lock.
- Added file service functionality.
- Fixed known issues.
- Optimized UI display.
- Resolved some browser compatibility issues.
- Automatic trigger configurations no longer restrict duplicate configurations.
- Optimized administrator tools, currently supporting:
Command | Purpose |
---|---|
python admin-tool.py reset_admin |
Reset administrator account |
python admin-tool.py reset_db_upgrade_seq |
Reset database update sequence |
python admin-tool.py clear_redis |
Clear Redis |
- Refactored subscription handling to enhance extensible subscription features. Currently supports subscribing to:
- MQTT
- Redis
- UI interface supports adding/removing tags for authorized links, automatic trigger configurations, and batch processes.
- Added DataKit data source and automatic discovery of local DataKit.
- Optimized built-in MySQL memory usage.
- Optimized worker unit memory usage.
- Adapted to MySQL 8.0.
- Removed unused packages.
- Optimized file upload handling.
- Removed Python package online queries (performance issues).
- Base image changed to
ubuntu:20.04
. - Python version changed to 3.8.
- PIP tool optimized.
- Added explanations for monkey patches in the script editor.
- Added script market functionality.
- Supports importing using
import __<script name>
instead ofimport <current script set name>__<script name>
. - Added script set cloning functionality.
- Automatic trigger configurations support manual execution.
- Optimized environment variable cache handling.
- Added function cache manager.
- Added function storage manager.
- Added system log viewer.
1.0.59 ~ 1.0.60 (2021-03-29)
DFF.RESP(...)
split into two functions:- Return data:
DFF.RSEP(data, status_code=None, content_type=None, headers=None, allow_304=False)
- Return file:
DFF.RESP_FILE(file_path, status_code=None, headers=None, allow_304=False, auto_delete=False, download=True)
- Return data:
- Other detail optimizations.
1.0.58 (2021-03-25)
- Optimized "File Tool".
1.0.57 (2021-03-25)
- Export script set functionality supports simultaneous export of authorization links, automatic trigger configurations, and batch processes.
- Optimized UI display.
- Added internationalization support.
- Some content provides an English interface.
- Added
reset-admin.py
script to reset the admin user password (usage:python reset-admin.py
). - Added PIP tool, supporting installation of Python packages via the UI interface.
- Added file tool, supporting basic operations such as browsing via the UI interface.
- Added more
DFF.*
functionalities.DFF.RSRC(...)
generates absolute paths for files under the resource catalog.DFF.RESP(...)
generates custom API return results.
- Exported function APIs support custom return formats/content (such as returning html pages, downloading files, etc.).
- Exported function APIs support file uploads.
- Adjusted the storage location for additional third-party Python packages (now
{installation directory}/data/resources/extra-python-packages
).
1.0.55 ~ 1.0.56 (2021-01-12)
- Worker added built-in libraries.
- Jinja2==2.11.2
- Mako==1.1.3
1.0.51 ~ 1.0.54 (2021-01-08)
- Reduced the number of displayed timelines for script package imports/exports and script library restore points (maximum 50).
1.0.47 ~ 1.0.50 (2021-01-07)
- Added MQTT data source support.
- Optimized automatic deployment scripts, added multiple options, see project README.md for details.
- Fixed known bugs.
1.0.44 (2020-12-10)
- Fixed the issue where unexpected retransmission of ID during modification of "authorization links" and "batch processes" caused request failures.
1.0.42 ~ 1.0.43 (2020-12-10)
- Added prompt labels for built-in script sets.
1.0.41 (2020-12-09)
- When configuring via environment variables, the
DFF_
prefix needs to be added. (For example: environment variableDFF_MYSQL_HOST
corresponds to the configurationMYSQL_HOST
.)
1.0.39 ~ 1.0.40 (2020-12-09)
- Fixed wording issues.
1.0.37 ~ 1.0.38 (2020-12-04)
- Reduced concurrency for queues 0 (system), 7 (debugging) to reduce basic resource consumption.
1.0.35 ~ 1.0.36 (2020-12-03)
- Image files in installation packages compressed using gzip.
1.0.32 ~ 1.0.34 (2020-12-03)
- When
MODE
value isprod
(default), no longer outputs specific MySQL and Redis operation SQL/commands. - Function timeout errors no longer output logs.
- Script logs/fault modules default off.
- Reduced stack trace data, default no longer outputs local variable information for each frame.
- Automatic deployment scripts automatically create logrotate configurations and support changing installation paths.
1.0.31 (2020-12-01)
- Fixed the bug where
DFF.FUNC(...)
failed to correctly pass variables.
1.0.30 (2020-12-01)
- DEBUG tasks now default to queue 7.
- Docker Stack file splits Workers into 4:
Queue Number | Purpose |
---|---|
0 |
System |
1 ~ 6 |
General |
7 |
Debugging |
8 ~ 9 |
CPU-intensive |
1.0.29 (2020-12-01)
- Added
hint
option to the@DFF.API
decorator, which can fill any content serializable as JSON, appearing in theextraConfigJSON
field of the function list.
1.0.28 (2020-11-29)
- Fixed the issue where the automatic backup function unexpectedly backed up migration data.
1.0.25 ~ 1.0.27 (2020-11-29)
- Added database automatic backup function.
- Fixed known issues.
1.0.24 (2020-11-27)
- "Authorization Links" support specifying custom IDs when added (but must start with
auln-
). - "Batch Processes" support specifying custom IDs when added (but must start with
bat-
).
1.0.23 (2020-11-27)
- Fixed known issues.
1.0.21 ~ 1.0.22 (2020-11-27)
- "Member Management" and "AccessKey Management" moved to the management section.
1.0.20 (2020-11-26)
- No longer uses
DataFlux.f(x)
naming, uniformly usesDataFlux Func
.
1.0.19 (2020-11-26)
- Deprecated
EXTRA_PYTHON_IMPORT_PATH_LIST
configuration, replaced withEXTRA_PYTHON_IMPORT_PATH
(supports only one address). - Fixed known faults.
1.0.18 (2020-11-20)
- Added queue pressure calculation module to protect queues (only for synchronous calls):
- Pressure <= 100: Normal processing.
- 100 < Pressure < 200: Randomly discard synchronous call requests with {Pressure - 100}% probability.
- Added queue pressure display to the "Management-Overview" page.
- Changed ID fields to
utf8mb4
encoding. - Added "Test" button to test connectivity for added data sources.
- Added
_DFF_HTTP_REQUEST
toscript context to save HTTP request details. - UI adjustments/optimizations.
- Fixed known issues.
Although this update modifies the database encoding, it does not provide an automatic migration function. Keeping the ID field as ASCII encoding will not affect system operation, but the search function on the list page will be unavailable.
1.0.15 ~ 1.0.17 (2020-11-06)
- Adjusted CD/CI configurations.
- Adjusted task queue naming and logic:
- Removed
default
queue. 0
as the fallback default queue for the entire system, mainly executing system-level tasks.1
as the default queue for function execution, functions execute by default in this queue.- Added
WORKER_QUEUE_ALIAS_MAP
parameter to support queue aliases, e.g.,cpu=9
indicates that when specifyingqueue='cpu'
, the function runs in queue9
.
- Removed
- During configuration loading, allows adding custom configurations with
CUSTOM_
prefix. Scripts can access configurations starting withCUSTOM_
. - Enhanced
DFF
object:- Added
DFF.STORE(key, scope=None)
access method. - Added
DFF.CACHE(key, scope=None)
access method. - Enhanced
DFF.SRC
, added the following functionalities:DFF.SRC.list()
lists data sources.DFF.SRC.save(...)
saves data sources.DFF.SRC.delete(...)
deletes data sources.
- Enhanced
DFF.ENV
, added the following functionalities:DFF.ENV.list()
lists environment variables.DFF.ENV.save(...)
saves environment variables.DFF.ENV.delete(...)
deletes environment variables.
- Added
DFF.CONFIG
to access custom configurations (CUSTOM_
prefixed items), including:DFF.CONFIG(...)
/DFF.get(...)
gets configurations.DFF.list()
lists configurations.
- Added
DFF.API(integration='autoRun', integration_config={...})
configuration, allowing functions to run automatically:integration_config={ 'crontab': '* * * * *' }
function executes periodically.integration_config={ 'onPublish': True }
function executes automatically after publishing.integration_config={ 'onLaunch': True }
function executes automatically upon system startup.
- Added
- Added statistical information view to the authorized link list page, including:
- Recent call counts.
- Recent response speeds (min, max, avg, mid).
- Recent response speed distribution (p75, p95, p99).
- Added
NSQ
data source. - Adjusted editor left tree list style.
- Added download button on script preview page (supports downloading drafts, published versions, and DIFFs).
- When functions use named parameters
**kwargs
, API examples, authorization link/batch processing configuration pages will display prompts for corresponding customizable parameters. - Upgraded WAT SDK to support configuring HTTP authentication headers.
- Fixed known issues, optimized code.
1.0.11 ~ 1.0.14 (2020-10-20)
- Fixed known issues.
1.0.10 (2020-10-20)
- Adjusted task queue naming and logic:
default
as the fallback default queue for the entire system, mainly executing system-level tasks.0
as the default queue for function execution, functions execute by default in this queue.1
~9
as extension queues, allowing functions to specify execution in these queues.- Functions executed during debugging in the editor are fixed to execute in queue
0
.
- Added
queue
parameter to the@DFF.API
decorator, allowing specification of the queue for function execution. - Added
api_timeout
parameter to the@DFF.API
decorator, allowing specification of the HTTP timeout time for functions as APIs. - Various default timeout times adjusted:
- Fixed timeout time of 60 seconds for debugging function execution in the editor.
- Default function execution timeout time is 30 seconds (can be specified between 1~3600 seconds via the
timeout
parameter of the@DFF.API
decorator). - Default function API timeout time is 10 seconds (can be specified between 1~180 seconds via the
api_timeout
parameter of the@DFF.API
decorator).
1.0.8 ~ 1.0.9 (2020-10-15)
- Adjusted function execution timeout, API interface call timeout, number of worker units, etc.
- Functions allow the use of
**kwargs
named parameters, and the system no longer rejects passing functions not in the parameter list.
1.0.7 (2020-10-13)
- Fixed the issue where additional Python packages could not be imported.
1.0.6 (2020-10-13)
- Added search functionality to the code editor.
- Increased Python keywords, DataFlux Func built-in function statement auto-completion in the code editor.
- Added "Learn Quickly" documentation link.
- Added
EXTRA_PYTHON_IMPORT_PATH_LIST
configuration to support additional Python import path lists. - Fixed known faults.
1.0.5 (2020-09-22)
- Added environment variables for resetting the administrator account (
RESET_ADMIN_USERNAME
,RESET_ADMIN_PASSWORD
).
1.0.3 ~ 1.0.4 (2020-09-20)
- Updated README content.
1.0.2 (2020-09-20)
- Updated automated deployment scripts.
1.0.1 (2020-09-20)
- Updated GitLab CI file.
- Updated database initialization file.
1.0.0 (2020-09-19)
DataFlux Func officially released.