Guillotina addon to split actions into smaller bits
Project description
Introduction
============
`guillotina_hive` is a task runner whose main goal is to provide a mechanism
to iterate through all the content on a database as quickly as possible.
To accomplish that end, hive integrates with container orchestrators(k8s or nomad)
to schedule jobs on and retrieve results from those jobs.
Installation
------------
With pip:
pip install guillotina_hive
Guillotina configuration
------------------------
Example here with json::
"applications": ["guillotina_hive"],
"hive": {
"default_image": None,
"default_namespace": "hive",
"orchestrator": "k8s",
"cluster_config": {},
"guillotina_default": {
"entrypoint": None,
"volumes": None,
"volumeMounts": None,
"envFrom": None
},
'quota': None
}
Settings explained
~~~~~~~~~~~~~~~~~~
- default_image: default image to use for jobs to run with
- default_namespace: cluster namespace to use
- orchestrator: k8s or nomad
- cluster_config: what to configure aioclustermanager with
- guillotina_default: default configuration for guillotina jobs
- quota: define quote on cluster namespace
Defining a job
--------------
We use decorators to provide tasks::
from guillotina_hive.decorators import hive_task
@hive_task(name='something')
async def something(arg1, arg2):
return foobar
You can also use application settings::
{
"calculate-numbers": "guillotina_hive.tests.tasks.calculate_numbers"
}
1.0.15 (2018-05-19)
-------------------
- Downgrade guillotina test requirements
1.0.14 (2018-05-19)
-------------------
- upgrade guillotina in reqs
1.0.13 (2018-05-19)
-------------------
- Do not log cancelled errors on init
[vangheem]
1.0.12 (2018-05-05)
-------------------
- job name should not be the id we give it
[vangheem]
- Be able to customize default container args
[vangheem]
1.0.11 (2018-04-27)
-------------------
- Updating aioclustermanager dependency
[ramon]
1.0.10 (2018-04-26)
-------------------
- Remove print
[vangheem]
1.0.9 (2018-04-26)
------------------
- apply builtin tasks setup request object url as well
[vangheem]
1.0.7 (2018-04-25)
------------------
- improve user login for apply tasks
[vangheem]
1.0.4 (2018-04-24)
------------------
- Fix working with decorators
[vangheem]
1.0.3 (2018-04-24)
=======
- fix to work with guillotina 3
[vangheem]
1.0.2 (27/04/2018)
------------------
- Updating aioclustermanager dependency
[ramon]
1.0.1 (9/4/2018)
----------------
- Adding manifest
[ramon]
1.0.0 (5/4/2018)
----------------
- Initial version
============
`guillotina_hive` is a task runner whose main goal is to provide a mechanism
to iterate through all the content on a database as quickly as possible.
To accomplish that end, hive integrates with container orchestrators(k8s or nomad)
to schedule jobs on and retrieve results from those jobs.
Installation
------------
With pip:
pip install guillotina_hive
Guillotina configuration
------------------------
Example here with json::
"applications": ["guillotina_hive"],
"hive": {
"default_image": None,
"default_namespace": "hive",
"orchestrator": "k8s",
"cluster_config": {},
"guillotina_default": {
"entrypoint": None,
"volumes": None,
"volumeMounts": None,
"envFrom": None
},
'quota': None
}
Settings explained
~~~~~~~~~~~~~~~~~~
- default_image: default image to use for jobs to run with
- default_namespace: cluster namespace to use
- orchestrator: k8s or nomad
- cluster_config: what to configure aioclustermanager with
- guillotina_default: default configuration for guillotina jobs
- quota: define quote on cluster namespace
Defining a job
--------------
We use decorators to provide tasks::
from guillotina_hive.decorators import hive_task
@hive_task(name='something')
async def something(arg1, arg2):
return foobar
You can also use application settings::
{
"calculate-numbers": "guillotina_hive.tests.tasks.calculate_numbers"
}
1.0.15 (2018-05-19)
-------------------
- Downgrade guillotina test requirements
1.0.14 (2018-05-19)
-------------------
- upgrade guillotina in reqs
1.0.13 (2018-05-19)
-------------------
- Do not log cancelled errors on init
[vangheem]
1.0.12 (2018-05-05)
-------------------
- job name should not be the id we give it
[vangheem]
- Be able to customize default container args
[vangheem]
1.0.11 (2018-04-27)
-------------------
- Updating aioclustermanager dependency
[ramon]
1.0.10 (2018-04-26)
-------------------
- Remove print
[vangheem]
1.0.9 (2018-04-26)
------------------
- apply builtin tasks setup request object url as well
[vangheem]
1.0.7 (2018-04-25)
------------------
- improve user login for apply tasks
[vangheem]
1.0.4 (2018-04-24)
------------------
- Fix working with decorators
[vangheem]
1.0.3 (2018-04-24)
=======
- fix to work with guillotina 3
[vangheem]
1.0.2 (27/04/2018)
------------------
- Updating aioclustermanager dependency
[ramon]
1.0.1 (9/4/2018)
----------------
- Adding manifest
[ramon]
1.0.0 (5/4/2018)
----------------
- Initial version
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
guillotina_hive-1.0.15.tar.gz
(16.4 kB
view hashes)
Built Distribution
Close
Hashes for guillotina_hive-1.0.15-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | bf34da63a23536339c4849e214b5958900723e2d80650fe1ea8e4b287c3ba9be |
|
MD5 | 284b5bb1e10d9f659ad4f4786462ec66 |
|
BLAKE2b-256 | ff8e7668c1f08e1310459a2cfa0056202805552651dee91bd49413bf16b303ee |