Can catchup be used to backfill failed dag runs in Airflow? - airflow

I have read that Airflows catchup feature applies to task instances that do not yet have a state - i.e. the scheduler will pick up any execution dates where the DAG has not yet ran (starting from the given start_date) - is this correct, and if so, does this mean catchup does not apply to failed DAG runs?
I am looking for a way to backfill any execution dates that failed, rather than not having ran at all.

Take a look at the backfill command options. You could use rerun-failed-tasks:
if set, the backfill will auto-rerun all the failed tasks for the backfill date range instead of throwing exceptions
Default: False
or reset-dagruns:
if set, the backfill will delete existing backfill-related DAG runs and start anew with fresh, running DAG runs
Default: False
Also, keep in mind this:
If reset_dag_run option is used, backfill will first prompt users whether airflow should clear all the previous dag_run and task_instances within the backfill date range. If rerun_failed_tasks is used, backfill will auto re-run the previous failed task instances within the backfill date range.
Before doing anything like the above, my suggestion is that you try it first with some dummy DAG or similiar.

Related

New task in DAG blocks further DAG executions

We have an ETL DAG which is executed daily. DAG and tasks have the following parameters:
catchup=False
max_active_runs=1
depends_on_past=True
When we add a new task, due to depends_on_past property, no new DAG runs get scheduled, as all previous states for new task are missing.
We would like to avoid having to run manual backfill or manually marking previous runs from UI as it can be easily forgotten, and we also have some dynamic DAGs where tasks get added automatically and halt future DAG executions.
Is there a way to automatically set past executions for new tasks as skipped by default, or some other approach that will allow future DAG runs to execute without human intervention?
We also considered creating a maintenance DAG that would insert missing task executions with skipped state, but would rather not go this route.
Are we missing something as the flow looks like a common thing to do?
Defined in Airflow documentation on BaseOperator:
depends_on_past (bool) – when set to true, task instances will run
sequentially and only if the previous instance has succeeded or has
been skipped. The task instance for the start_date is allowed to run.
As long as there exists a previous instance of the task, if that previous instance is not in the success state, the current instance of the task cannot run.
When adding a task to a DAG with existing dagrun, Airflow will create the missing task instances in the None state for all dagruns. Unfortunately, it is not possible to set the default state of task instances.
I do not believe there is a way to allow future task instances of a DAG with existing dagruns to run without human intervention. Personally, for depends_on_past enabled tasks, I will mark the previous task instance as success either through the CLI or the Airflow UI.
Looks like there is an Github Issue describing exactly what you are experiencing! Feel free to bump this PR or take a stab at it if you would like.
A hacky solution is to set depends_on_past to False as max_active_runs=1 will implicitly guarantee the same behavior. As of the current Airflow version, the scheduler orders both dag runs and task instances by execution date before running them (checked 1.10.x but also 2.0)
Another difference is that next execution will be scheduled even if previous fails. We solved this by retrying unlimited times (setting a ver large number), and alert if retry number is larger than some value.

Airflow DAG Multiple Runs

I have a DAG that I want to run multiple times after each successful completion. For an example I want to run it 10 times and stop. Is there a way to accomplish this? I tried looking into scheduling with CRON but it doesn't seem clean nor triggering the DAG via UI multiple times doesn't work (runs in parallel).
I found a solution to my use case. It incorporated using depends_on_past=True (mentioned by #Hitesh Gupta) and setting your airflow.cfg file below:
# The maximum number of active DAG runs per DAG
max_active_runs_per_dag = 1
This allowed us to only have one active DAG run at a time and also to not continue the next DAG run if there were failure in the previous run. This is for Airflow version 1.10.1 that I tested on.
You can, in addition to supplying a start_date, provide your DAG an end_date
Quoting the docstring
:param start_date: The timestamp from which the scheduler will attempt
to backfill
:type start_date: datetime.datetime
:param end_date: A date beyond which your DAG won't run, leave to None for open ended scheduling
:type end_date: datetime.datetime
While unrelated, also have a look at following scheduler settings in airflow.cfg as mentioned in this article
run_duration
num_runs
UPDATE-1
In his article Use apache airflow to run task exactly once, #Andreas P has described a clever technique, which I believe can be adapted to your use-case. While even that won't be a very-tidy solution, it would at-least allow you to specify beforehand the number of runs (integer) for DAG instead of end_date.
Alternatively (assuming you implement the above approach) rather than rather than baking this skipping-dag-after max-runs functionality within each DAG, you can create a separate orchestrator DAG that disables a given DAG after its max runs have passed.
You have to set property depends_on_past. This is set under DAG's default arguments section and it refers to previous instance dag instance. This is fix your problem.

Is it possible to have airflow backfill and scheduling at the same time?

I am in a situation where I have started getting some data scheduled daily at a certain time and I have to create ETL for that data.
Meanwhile, when I am still creating the DAGs for scheduling the tasks in Airflow. The data keeps on arriving daily. So when I will start running my DAGs from today I want to schedule it daily and also wants to backfill all the data from past days which I missed while I was creating DAGs.
I know that if I put start_date as the date from which the data started arriving airflow will start backfilling from that date, but wouldn't in that case, my DAGs will always be behind of current day? How can I achieve backfilling and scheduling at the same time? Do I need to create separate DAGs/tasks for backfill and scheduling?
There are several things you need to consider.
1. Is your daily data independent or the next run is dependent on the previous run?
If the data is dependent on previous state you can run backfill in Airflow.
How backfilling works in Airflow ?
Airflow gives you the facility to run past DAGs. The process of running past DAGs is called Backfill. The process of Backfill actually let Airflow forset some status of all DAGs since it’s inception.
I know that if I put start_date as the date from which the data
started arriving airflow will start backfilling from that date, but
wouldn't in that case, my DAGs will always be behind of current day?
Yes setting a past start_date is the correct way of backfilling in airflow.
No, If you use celery executer, the jobs will be running in parallel and it will eventually catch up to the current day , obviously depending upon your execution duration.
How can I achieve backfilling and scheduling at the same time? Do I
need to create separate DAGs/tasks for backfill and scheduling?
You do not need to do anything extra to achieve scheduling and backfilling at the same time, Airflow will take care of both depending on your start_date
Finally , If this activity is going to be one time task I recommend , you process your data(manually) offline to airflow , this will give you more control over the execution.
and then either mark the backfilled tasks as succeed or below
Run an airflow backfill command like this: airflow backfill -m -s "2016-12-10 12:00" -e "2016-12-10 14:00" users_etl.
This command will create task instances for all schedule from 12:00 PM to 02:00 PM and mark it as success without executing the task at all. Ensure that you set your depends_on_past config to False, it will make this process a lot faster. When you’re done with it, set it back to True.
Or
Even simpler set the start_date to current date

how to get airflow DAG execution date using command line?

In order for me to get the dag_state, I run the following LCI command:
airflow dag_state example_bash_operator '12-12T16:04:46.960661+00:00'
The trouble is - I have to explicitly pass the exact date-time (i.e. execution_date) to this command.
When I run airflow list_dags I only get a listing of DAG's but not their execution dates.
Is there a way to obtain the exact date time (i.e. -> '12-12T16:04:46.960661+00:00')
for a given dag, using command line CLI?
There's a conceptual issue here. Dags are objects that have schedules, not execution dates. When the schedule is due, DagRuns are created for that Dag with the appropriate execution_date.
So you can ask for the state of a DagRun using the CLI and providing the execution_date, because execution dates (almost uniquely) map to a specific DagRun. Almost uniquely because in practice you can trigger two DagRuns with the same execution_date, but that's an unusual scenario.
But if you ask for the execution_date of a Dag, what do you really want to know? The execution_date of the last recently created DagRun? The list of execution_dates for the currently running DagRuns?
You can check list_dag_runsdag_id CLI command and see if yon can filter it to your needs.

Apache Airflow tasks are stuck in a 'up_for_retry' state

I've been setting up an airflow cluster on our system and previously it has been working. I'm not sure what I may have done to change this.
I have a DAG which I want to run on a schedule. To make sure it's working I'd also like to trigger it manually. Neither of these seem to be working at the moment and no logs seem to be being written for the task instances. The only logs available are the airflow scheduler logs which generally look healthy.
I am just constantly met with this message:
Task is not ready for retry yet but will be retried automatically. Current date is 2018-12-12T11:34:46.978355+00:00 and task will be retried at 2018-12-12T11:35:08.093313+00:00.
However, if I wait a little the exact same message is presented again except the times have moved forward a little. Therefore, it seems the task is never actually being retried.
I am using a LocalExecutor and the task is an SSHOperator. Simplified code below. All it does is ssh onto another machine and start a bunch of application with a pre-determined directory structure.:
DB_INFO_FILE = 'info.json'
START_SCRIPT = '/bin/start.sh'
TIME_IN_PAST = timezone.convert_to_utc(datetime.today() -
timedelta(days=1))
DEFAULT_ARGS = {
'owner': 'airflow',
'depends_on_past': False,
'start_date': TIME_IN_PAST,
'email': ['some_email#blah.com'],
'email_on_failure': False,
'email_on_retry': False,
'retries': 1,
'retry_delay': timedelta(minutes=1),
}
def _extract_instance_id(instance_string):
return re.findall(r'\d+', instance_string)[0]
def _read_file_as_json(file_name):
with open(file_name) as open_file:
return json.load(open_file)
DB_INFO = _read_file_as_json(DB_INFO_FILE)
CONFIG_CLIENT = ConfigDbClient(**DB_INFO)
APP_DIRS = CONFIG_CLIENT.get_values('%my-app-info%')
INSTANCE_START_SCRIPT_PATHS = {
_extract_instance_id(instance_string): directory+START_SCRIPT
for instance_string, directory in APP_DIRS.items()
}
# Create an ssh hook which refers to pre-existing connection information
# setup and stored by airflow
SSH_HOOK = SSHHook(ssh_conn_id='my-conn-id')
# Create a DAG object to add tasks to
DAG = DAG('my-dag-id',
default_args=DEFAULT_ARGS)
# Create a task for each app instance.
for instance_id, start_script in INSTANCE_START_SCRIPT_PATHS.items():
task = SSHOperator(
task_id='run-script-{0}'.format(instance_id),
command='bash {0}'.format(start_script),
ssh_hook=SSH_HOOK,
dag=DAG)
It works when I run the tasks individually, via the command line but not via the UI. It seems I can run tasks but I simply cannot trigger a DAG to run. I've tried many combinations of start_date s and interval schedules just to sanity check also.
Any ideas?
And yes, I am aware this question has been asked before and I have looked at all of them but not of the solutions have helped me.
Oh. Your start_date is changing at the same rate or faster than the schedule interval period.
Here's what the scheduler is seeing every couple of seconds:
start_date: 2018-12-11T12:12:12.000Z # E.G. IFF now is 2018-12-12T12:12:12.000Z, a day ago
schedule_interval: timedelta(days=1) # The default
Here's what the scheduler needs for a DAG to run: The last time a run occurred was more than one schedule interval ago. If no scheduled run has occurred, the first scheduled run could start now if one full schedule interval has passed since the start_date as that is the earliest allowable date for execution_date. In which case the dag_run with the execution_date set to the start of that interval period should be created. Then task_instances can be created for any tasks in the DAG whose dependencies are met as long as the task_instance execution_date is after the start_date of the DAG (this is not stored on the dag_run object but recomputed by loading the DAG file just while inspecting the dag's state).
So it's not getting scheduled automatically for the reason that the start date keeps changing just as the interval is satisfied. However if it were -2d at least one run would get scheduled and then any further runs would have to wait until it's 1d after that to be scheduled. It's easier though if you just set a fixed datetime on your start_date.
But what about those odd retries on your manual runs…
You did start a manual run or two. These runs took the current time as the execution_date unless you specified something else. This should be after the start_date, at least until tomorrow, which should clear them to run. But then it seems in your logs you're seeing that they're failing and being marked for retry and also not decrementing your retries. I'm not sure why that would be but could it be that something isn't right with the SSHOperator.
Did you install airflow with the [ssh] extra so that SSHOperator's dependencies are met (specifically paramiko and sshtunnel) on both the webserver and scheduler? One of them is working because I assume it's parsing and showing up in the UI based on being added to the DB.
What do you get if you execute:
airflow test my-dag-id run-script-an_instance_id 2018-12-12T12:12:12
You know that the scheduler and webserver are looping over refilling their DAG bag and so rerunning this DAG file a few 1000 times a day, reloading that json (it's local access, so similar to importing a module), and recreating that SSHHook with a DB lookup. I don't see anything fancy done setting up this hook, why not just remove ssh_hook from the SSHOperator and replace it with ssh_conn_id='my-conn-id' so it can be created once at execution time?
I doubt that's the issue that's causing the retries that just roll forward though.
I had a task stuck in up_for_retry for almost 24 hours before I noticed it, and it had nothing to do with the start_date, end_date, or any other classic beginner's problem.
I resorted to reading the source code, and found that Airflow treats up_for_retry tasks differently if they are part of a backfill DAG run.
So I connected to the metadata DB and changed backfill_ to scheduled__ in the dag_run row corresponding to the stuck task. Airflow started running the stuck task immediately.

Resources