Table of Contents
Sequent provides programmer with interface to create flow. Sequent is useful when there are multiple batch processes that should link together in processing flow.
For example, programs A1, A2 extract data from database, program B process the extracted files, and programs C1 and C2 transmit process information to remote client.
Sequent can be used to create the flow where A1, A2 would run concurrently, then program B would run. When B is successful, programs C1 and C2 would run.
Sequent uses Eventor (https://github.com/Acrisel/eventor) as its underline infrastructure. With that, it gains Eventor's recovery capabilities. When running in recovery, successful steps can be skipped.
Considering the above example, this can be useful when C2 transmission program fails. Recovery run would only execute it with out redoing potentially expensive work that was already completed (A1, A2, B, and C1)
It would be easier to show an example. In this example, Step s2 would run after Step s1 would loop twice. When they are done, Step s3 will run. This flow would be repeat twice.
import sequent as seq import logging def prog(progname, success=True): logger = logging.getLogger(os.getenv("SEQUENT_LOGGER_NAME")) logger.info("doing what {} is doing".format(progname)) if not success: raise Exception("{} failed".format(progname)) return progname myflow = seq.Sequent(config={'sleep_between_loops': 0.05, 'LOGGING': {'logging_level':logging.DEBUG}}) s1 = myflow.add_step('s1', repeats=[1,2]) s11 = s1.add_step('s11', repeats=[1,2,]) s111 = s11.add_step('s111', func=prog, kwargs={'progname': 'prog1'}) s112 = s11.add_step('s112', func=prog, kwargs={'progname': 'prog2'}, requires=( (s111, seq.STEP_SUCCESS), )) s12 = s1.add_step('s12', func=prog, kwargs={'progname': 'prog3'}, require=s( (s11, seq.STEP_SUCCESS), )) s2 = myflow.add_step('s2', func=prog, kwargs={'progname': 'prog4'}, requires=((s1, seq.STEP_SUCCESS),)) myflow.run()
The above example with provide the following log output. Note more detailed logging activities was stripped off. Only actual shows actual program activity is shown.
[ 2016-12-07 11:14:29,761 ][ INFO ][ Eventor store file: /sequent/example/runly00.run.db ] ... [ 2016-12-07 11:14:29,962 ][ INFO ][ doing what prog1 is doing ] ... [ 2016-12-07 11:14:30,124 ][ INFO ][ doing what prog2 is doing ] ... [ 2016-12-07 11:14:30,358 ][ INFO ][ doing what prog1 is doing ] ... [ 2016-12-07 11:14:30,587 ][ INFO ][ doing what prog2 is doing ] ... [ 2016-12-07 11:14:30,908 ][ INFO ][ doing what prog3 is doing ] ... [ 2016-12-07 11:14:31,234 ][ INFO ][ doing what prog1 is doing ] ... [ 2016-12-07 11:14:31,407 ][ INFO ][ doing what prog2 is doing ] ... [ 2016-12-07 11:14:31,657 ][ INFO ][ doing what prog1 is doing ] ... [ 2016-12-07 11:14:31,894 ][ INFO ][ doing what prog2 is doing ] ... [ 2016-12-07 11:14:32,240 ][ INFO ][ doing what prog3 is doing ] ... [ 2016-12-07 11:14:32,565 ][ INFO ][ doing what prog4 is doing ] ... [ 2016-12-07 11:14:32,713 ][ INFO ][ Processing finished with: success ]
Flow diagram:
+--S1----------------------+ | | | +--S11----------+ | | | | | | | S111 -> S112 | -> S12 | -> S2 | +---------------+ | +--------------------------+For simplicity, code definition of prog (line 6) serves as reusable activity for all the steps in this example.
A Sequent object is defined (line 12) to host myflow. By default, Sequent's Eventor loops on events and steps. By defaults it sleeps one second between loops. Here 'sleep_between_loops' changes this setting to 0.05 seconds.
myflow contains two steps, s1 and s2. s1 is a container step that would repeat twice (defined on line 15). s2 is a processing step (defined on line 26).
s1 contains two steps. s11 (line 17) is container step and s12 is a processing step.
s11 contains two processing steps s111 and s112 (lines 19-20).
Finally, on line 29 the flow is executed using myflow().
logger is set with in step program (line 5) to direct step logging into its dedicated log.
Sequent signature in its most simplistic format:
Sequent(name='', store='', run_mode=SEQ.SEQUENT_RESTART, recovery_run=None, config={}, config_tag='')
Sequent, when instantiated, provides interface to build program flow. When called upon, Sequent steps are translated to Eventor steps and Step's requires are translated to Eventor's Events and Steps' triggers.
Sequent instantiation arguments are the same as Eventor's.
name: string id for Sequent object initiated.
- store: path to file that would store runnable (sqlite) information; if ':memory:' is used, in-memory temporary
storage will be created. If not provided, calling module path and name will be used with db extension instead of 'py'.
- run_mode: can be either RUN_RESTART (default) or RUN_RECOVER; in restart, new instance or the run
will be created. In recovery,
- recovery_run: if RUN_RECOVER is used, recovery_run will indicate specific instance of previously recovery
run that would be executed.If not provided, latest run would be used.
config: keyword dictionary of default configurations. Available keywords and their default values:
Name Default Value Description workdir /tmp place to create necessry artifacts (not in use) logdir /var/log/eventor place to create debug and error log files task_construct 'process' method to use for execution of steps max_concurrent 1 maximum concurrent processing, if value <1, no limit will be pose stop_on_exception True if an exception occurs in a step, stop all processes. If True, new processes will not start. But running processes will be permitted to finish sleep_between_loops 1 seconds to sleep between iteration of checking triggers and tasks config_tag: key with in config where Sequent configuration starts.
add_event(require=None)
- requires: logical expression 'sqlalchemy' style to automatically raise this expression.
syntax:
requires : (requires, requires, ...) | or_(requires, requires, ...) | event
- if expression is of the first style, logical and will apply.
- the second expression will apply logical or.
- the basic atom in expression is even which is the product of add_event.
Event object to use are require in add_step.
add_step(name, func, args=(), kwargs={}, hosts=[], requires={}, delay=0, acquires=[], releases=None, recovery={}, config={})
name: string unique id for step
func: callable object that would be call at time if step execution
args: tuple of values that will be passed to func at calling
kwargs: keywords arguments that will be passed to func at calling
- hosts: list of hosts step should run on. If not provided, localhost will be used.
if
requires: mapping of step statuses such that when set of events, added step will be launched:
status description STEP_READY set when task is ready to run (triggered) STEP_ACTIVE set when task is running STEP_SUCCESS set when task is successful STEP_FAILURE set when task fails STEP_COMPLETE stands for success or failure of task
- delay: seconds to wait before executing step once is requires are available. Actual execution
may be delayed further if resources needs to be acquired.
acquires: list of tuples of resource pool and amount of resources to acquire before starting.
- releases: list of tuples of resources pool and amount of resources to release once completed.
If None, defaults to acquires. If set to empty list, none of the acquired resources would be released.
recovery: mapping of state status to how step should be handled in recovery:
status default description STEP_READY STP_RERUN if in recovery and previous status is ready, rerun STEP_ACTIVE STP_RERUN if in recovery and previous status is active, rerun STEP_FAILURE STP_RERUN if in recovery and previous status is failure, rerun STEP_SUCCESS STP_SKIP if in recovery and previous status is success, skip config: keywords mapping overrides for step configuration.
name default description stop_on_exception True stop flow if step ends with Exception
Step object to use in add_assoc method.
run(max_loops=-1)
when calling run method, information is built and loops evaluating events and task starts are executed. In each loop events are raised and tasks are performed. max_loops parameters allows control of how many loops to execute.
In simple example, myflow.run() engage Sequent's run() method.
- max_loops: max_loops: number of loops to run. If positive, limits number of loops.
defaults to negative, which would run loops until there are no events to raise and no task to run.
If there was a failure that was not followed by event triggered, result will be False.
Sequent can operate Steps on distributed environment. A step can be associated with hosts using hosts argument in add_step. Sequent uses SSH to submit steps to remote host. This means that cluster needs to be configured with SSH keys. To set up the environment for Sequent distributed operation:
- Host from which Sequent program would be initiated, should be able to SSH to participating hosts without only using keys.
- SSH authorized_keys on each target host should has proper command to initiate the right operation environment. This may include activating the correct virtualenv.
- Optionally, set SSH backdoor to originated host. In the future Sequent may use this backdoor, as callback.
- Software needs to be uniformly installed on all participating machines.
- Sequent must be initiated with database configuration that is accessible from all participating hosts. Sequent and its remote agents would use that database to share operation information. The database user needs to have permissions to create schema (if the associated schema is not created.) It also needs to have create table permissions.
Anything passed to Sequent, predominately with add_step, needs to be importable. For example in simple example:
import example_progs as example s111 = s11.add_step('s111', func=example.prog, kwargs={'progname': 'prog1'})
Recovery allows rerun of a program in a way that it will skip successful steps. To use recovery, store mast be physical (cannot use in-memory).
According to step recovery setup, when in recovery, step may be skipped or rerun. By default, only success statuses are skipped.
Here is an example for recovery program and run.
import sequent as sqnt import logging appname = os.path.basename(__file__) logger = logging.getLogger(appname) def prog(flow, progname, step_to_fail=None, iteration_to_fail=''): logger = logging.getLogger(os.getenv("SEQUENT_LOGGER_NAME")) step_name = flow.get_step_name() step_sequence = flow.get_step_sequence() logger.info("doing what {} is doing (}/{})".format(progname, step_name, step_sequence)) if step_to_fail == step_name and step_sequence== iteration_to_fail: raise Exception("{} failed ({}/{})".format(progname, step_name, step_sequence)) return progname def build_flow(run_mode = sqnt.RUN_RESTART, run_id=None, step_to_fail=None, iteration_to_fail=''): myflow = sqnt.Sequent(name=appname, run_mode=run_mode, run_id=run_id, config={'sleep_between_loops': 0.05,}, ) s1 = myflow.add_step('s1', repeats=[1,2]) s11 = s1.add_step('s11', repeats=[1,2,]) s111 = s11.add_step('s111', func=prog, kwargs={'flow': myflow, 'progname': 'prog1', 'step_to_fail':step_to_fail, 'iteration_to_fail':iteration_to_fail,}) s112 = s11.add_step('s112', func=prog, kwargs={'flow': myflow, 'progname': 'prog2', 'step_to_fail':step_to_fail, 'iteration_to_fail':iteration_to_fail,}, requires=((s111, sqnt.STEP_SUCEESS),)) s12 = s1.add_step('s12', func=prog, kwargs={'flow': myflow, 'progname': 'prog3', 'step_to_fail':step_to_fail, 'iteration_to_fail':iteration_to_fail,}, requires=((s11, sqnt.STEP_SUCEESS),)) s2 = myflow.add_step('s2', func=prog, kwargs={'flow': myflow, 'progname': 'prog4', 'step_to_fail':step_to_fail, 'iteration_to_fail':iteration_to_fail,}, requires=((s1, sqnt.STEP_SUCEESS),)) return myflow # creating flow simulating failure myflow = build_flow(step_to_fail='s1_s11_s111', iteration_to_fail='1.2.2') myflow.run() run_id = myflow.run_id # creating recovery flow myflow = build_flow(run_mode=RUN_RECOVER, run_id=run_id) myflow.run()
[ 2016-12-07 14:49:24,437 ][ INFO ][ Eventor store file: /sequent/example/runly04.run.db ] ... [ 2016-12-07 14:49:24,645 ][ INFO ][ doing what prog1 is doing (s1_s11_s111/1.1.1) ] ... [ 2016-12-07 14:49:24,805 ][ INFO ][ doing what prog2 is doing (s1_s11_s112/1.1.1) ] ... [ 2016-12-07 14:49:25,047 ][ INFO ][ doing what prog1 is doing (s1_s11_s111/1.1.2) ] ... [ 2016-12-07 14:49:25,272 ][ INFO ][ doing what prog2 is doing (s1_s11_s112/1.1.2) ] ... [ 2016-12-07 14:49:25,587 ][ INFO ][ doing what prog3 is doing (s1_s12/1.1) ] ... [ 2016-12-07 14:49:25,909 ][ INFO ][ doing what prog1 is doing (s1_s11_s111/1.2.1) ] ... [ 2016-12-07 14:49:26,073 ][ INFO ][ doing what prog2 is doing (s1_s11_s112/1.2.1) ] ... [ 2016-12-07 14:49:26,321 ][ INFO ][ doing what prog1 is doing (s1_s11_s111/1.2.2) ] [ 2016-12-07 14:49:26,323 ][ INFO ][ [ Step s1_s11_s111/1.2.2 ] Completed, status: TaskStatus.failure ] [ 2016-12-07 14:49:26,397 ][ ERROR ][ Exception in run_action: <Task(id='15', step_id='s1_s11_s111', sequence='1.2.2', recovery='0', pid='10276', status='TaskStatus.failure', created='2016-12-07 20:49:26.300030', updated='2016-12-07 20:49:26.311884')> ] [ 2016-12-07 14:49:26,397 ][ ERROR ][ Exception('prog1 failed (s1_s11_s111/1.2.2)',) ] [ 2016-12-07 14:49:26,397 ][ ERROR ][ File "/eventor/eventor/main.py", line 63, in task_wrapper result=step(seq_path=task.sequence) File "/eventor/eventor/step.py", line 82, in __call__ result=func(*func_args, **func_kwargs) File "/sequent/example/runly04.py", line 34, in prog raise Exception("%s failed (%s/%s)" % (progname, step_name, step_sequence)) ] [ 2016-12-07 14:49:26,397 ][ INFO ][ Stopping running processes ] [ 2016-12-07 14:49:26,401 ][ INFO ][ Processing finished with: failure ] [ 2016-12-07 14:49:26,404 ][ INFO ][ Eventor store file: /sequent/example/runly04.run.db ] ... [ 2016-12-07 14:49:27,921 ][ INFO ][ doing what prog1 is doing (s1_s11_s111/1.2.2) ] ... [ 2016-12-07 14:49:28,159 ][ INFO ][ doing what prog2 is doing (s1_s11_s112/1.2.2) ] ... [ 2016-12-07 14:49:28,494 ][ INFO ][ doing what prog3 is doing (s1_s12/1.2) ] ... [ 2016-12-07 14:49:28,844 ][ INFO ][ doing what prog4 is doing (s2/1) ] [ 2016-12-07 14:49:28,845 ][ INFO ][ [ Step s2/1 ] Completed, status: TaskStatus.success ] [ 2016-12-07 14:49:29,002 ][ INFO ][ Processing finished with: success ]
The function build_flow (code line 14) build a Sequent flow similarly to simple example above. Since no specific store is provided in Sequent instantiation, a default runner store is assigned (code line 15). In this build, steps will use default recovery directives whereby successful steps are skipped.
The first build and run is done in lines 42-43. In this run, a parameter is passed to cause step s111 in its fourth iteration to fail. As a result, flow fails. Output lines 1-29 is associated with the first run.
The second build and run is then initiated. In this run, parameter is set to a value that would pass step s111 and run mode is set to recovery (code lines 45-46). Eventor skips successful steps and start executing from failed steps onwards. Output lines 30-40 reflects successful second run.
Note that the second run required a run_id of the run that is reactivated. run_id is fetched from its corresponding attribute in Sequent Objects.
For prog to know when to default, it uses the following methods flow.get_step_name() and flow.get_step_sequence() (lines 7-8). Those Sequent methods allow access to Eventor step attributes. Another way to access these attributes is via os.environ:
name = os.getenv('SEQUENT_STEP_NAME') sequence = os.getenv('SEQUENT_STEP_SEQUENCE') recovery = os.getenv('SEQUENT_STEP_RECOVERY') logger_name = os.getenv('SEQUENT_LOGGER_NAME')
add_step allows association of step with resources. If acquires argument is provided, before step starts, Eventor will attempt to reserve resources. Step will be executed only when resources are secured.
When release argument is provided, resources resources listed as its value will be released when step is done. If release is None, whatever resources stated by acquires would be released. If the empty list is set as value, no resource would be released.
To use resources, program to use Resource and ResourcePool from acris.virtual_resource_pool. Example for such definitions are below.
import sequent as sqnt from acris import virtual_resource_pool as vrp class Resources1(vrp.Resource): pass class Resources2(vrp.Resource): pass rp1 = vrp.ResourcePool('RP1', resource_cls=Resources1, policy={'resource_limit': 2, }).load() rp2 = vrp.ResourcePool('RP2', resource_cls=Resources2, policy={'resource_limit': 2, }).load() myflow = sqnt.Sequent(config={'sleep_between_loops': 0.05,}, ) s1 = myflow.add_step('s1', repeats=[1,2], acquires=[(rp1, 2), ])
Sequent github project (https://github.com/Acrisel/sequent) has additional examples with more complicated flows.