Scheduler
Azure Scheduler allows you to invoke actions that call HTTP/S endpoints or post messages to a storage queue on any schedule. In Scheduler, you can create jobs that reliably call services either inside or outside of Windows Azure and run those jobs right away, on a regular or irregular schedule, or set them to run at a future date.
If you have any feedback on Azure Scheduler, we’d love to hear it.
-
For Scheduler REST API, Update Job, should allow setting "schedule=null"
The current update job REST API (with PATCH) in Scheduler does not allow recurrence frequency changing from week to hour, when the "advanced schedule" is set. That is because "advanced schedule" is only meaningful for weekly schedule. I tried to set "schedule=null", but it did not work. This means I have to use a GET to get the Job, and then a PUT to update the Job in this case. I think a better idea would be letting the user to update schedule to be null, and letting them change recurrence frequency from week to hour
8 votes -
Append to job history
If my backend work takes longer than the http request time (asynchronous) or for queue based requests, I would like to have an API to append status to the history of the job to indicate that the actual backend work is completed.
7 votes -
Collection of generic task to be scheduled
There are a lot of generic tasks that we may have to schedule. For instance shutting down/ up some VMs or other resources...
Why not having the choice of selecting in a built-in collection of tasks as an alternative to select an url?
6 votesThis is a great idea. Please include a list of the top most generic tasks you would like to see.
-
Improve "programmatic" scheduling scenario
I'll combine multiple ideas for a single scenario.
Currently scheduler is designed towards manual scheduling scenarios. For example, if you are administering certain service and need to schedule something from time to time, it's easy to use it this way.
However, for scenarios where jobs are created as a result of user actions (not admin's actions), current scheduler falls short.
For example, consider a more or less modern website. When a user is registering at the website you need to send a welcome email - that's ok, for this case you don't actually need a scheduler as this is an…
5 votesMr. Frostlander,
Thank you for sharing your feedback.
In regards to programmatic ISV scenarios we had designed the job collections with intention of mapping either an ISV user or application representation to a job collection so that the ISV entity can have multiple jobs and they can be managed together. In your scenario you have not created that mapping. If we were to increase the job collection limit for a particular plan what would a reasonable limit be?
In regards to authentication the Scheduler service has a new REST interface behind the Azure Resource Manager (ARM) which allows you to use OAuth tokens. Would that work for you?Regards,
Kevin Lam -
Version control in Scheduling Services
Currently Version control is not enabled in Scheduler.
Code checkin and check out and team coding will help4 votesToday you can save the Scheduler definition as JSON via PowerShell and the API and check that into source control. If there is something that you are looking for in particular in regards to source control would you mind elaborating?
Thanks,
Kevin -
add a method to the portal for downloading/viewing the exact message placed in a queue by a job
This would aid debugging services that pull jobs from the queue placed by the scheduler
4 votesWe will look into including a sample preview of the message in the portal.
-
Add Refresh in Scheduler Job Collection blade
Reproduce step.
- Scheduler Job Collection blade > Add
- Add a job and click OK.
- Return to Scheduler Job Collection blade, but Monitoring section is not refreshed job list.
So, it would be good to add Refresh button on the Scheduler Job Collection blade.
4 votes -
hybrid worker runbook drainstop for patching/reboot
It would be nice if there were a way to "drain stop" a hybrid worker so that the node could complete any running runbook and not start any new ones.
If an Azure runbook is running on a hybrid worker and the server is rebooted, it will likely re-start the runbook on another node which is good, however if the killed job had already completed a step in the runbook which made a change/uploaded data, etc, and the job is restarted, duplication can occur.3 votes -
Allow dynamic parameters
The idea is to send with my Http Get request some dynamic parameters like the last time this particular job ran, so I could add to my Uri something like "http://myapi.domain/CallFromScheduler?lasttime=@JobLastRun@".
Others parameters could be added like:
1) @JobLastRun@: UTC Datetime of last run
2) @JobStatusLastRun@: Status of last run
3) @JobNextRun@: UTC Datetime of next run
4) @JobLastBody@: Body of response of previous run3 votesThanks for you feedback.
-Kevin
-
Run error config now
Along with "Run now", it would be good to place "Run error config now" for test purpose.
Currently, I make an error (server down) and click "Run now", then wait for several retry counts in order to confirm my error config settings correctly.
3 votesGreat idea. Thanks.
-
This needs job completion to trigger other jobs and allow job dependencies
We use CA 7 for job scheduling, which offers triggering of other jobs based on successful completion. Also jobs that have dependencies on other jobs. I guess this is more of an automated workload processing system, and not a simple scheduler. Are there plans to make Scheduler this powerful?
3 votes -
Clone button could be helpful
It would be good to place Clone button on Scheduler Job blade like logic app for making similar jobs faster.
3 votes -
Add TCP connection
It would be good to support TCP connection for IoT devices and TCP-based network routing/proxy devices. At first, only reachability checkup is helpful.
2 votes -
Increase max Message Body Size
I would like to be able to send a greater message body for my Jobs.
I think at least 64Kb message body size should be supported so it can be used with the maximum payload size of a storage queue message.
2 votesThe challenge with upping the limit of the max message size (which is 8K today) is that it increases storage on our backend as we would have to keep those large messages around to be able to send it on each occurrence a job. Referencing a blob is a nice idea for being able to mitigate that. I will add that to our backlog.
Thanks for your feedback.-Kevin
-
Add the ability to re-run schedules
Allow scheduler to run something from N-days ago, passing in the N-1 date as appropriate
1 voteHi Dan,
I’m not quite sure I fully understand your scenario. Are you asking for us to allow you to take a completed job and change the schedule so that it will start running again on the new schedule? Or are you asking for a job to be run immediately (run now)?Thanks,
Kevin -
Scheduler Error Policy following Recurrence policy logic
On this page (https://azure.microsoft.com/en-us/documentation/articles/scheduler-concepts-terms/) under retryPolicy it says that the default retry policy is 5 times with an interval of 1 minute. This isn't the case; it is 4 times with an interval of 30 seconds. Secondly, when I change it to a custom retry policy with retries every minute, it tells me that this is more often than the 1 hour retry policy allowed by this collection plan (because i'm on the free scheduler)
Every other piece of documentation says that that 1 hour is how often the scheduler can fire a job, and says nothing about…
1 voteThank you for calling this out. I have put in a request to update the documentation to correctly indicate the default values.
The FREE plan uses the min recurrence configuration setting (1 hour) as its lower bound for configuring the retry policy. The documentation will be updated to make this clearer as well. -
scheduler: job collection auth
Hi,
The Azure Scheduler allow us to use Basic Auth.
However, we have to defined credentials for each Job.Because I'm planing, for my company, to use it for a large number of jobs / job collections. I would like to configure the Basic Auth for each Job Collection instead of Jobs directly.
Ex : If we need for any security reason to change our credentials, we won't need to update hundreds of jobs, but only few job collection.
Is it possible to override the feature to get the Job Basic Auth if defined or Job collection instead or nothing…
1 voteThank you for submitting this idea! This has been something that we have been considering but were waiting for customer feedback. In regards to your last question about overriding the feature could you provide more clarity as to what you are looking for?
Thanks,
Kevin -
Notification should be displayed when you save job with past "Start on" time
If you change recurrence from "Recurring" to "Once" and set a job to "Start on" past time, this job runs immediately after saving. This behavior is a little bit confusing to users who expect the job should not start. It’s very helpful if a caution appears on the portal when users are going to save jobs with past time.
1 vote -
For the scheduled tasks product, just place a dropdown with time zone so we know what time we are entering
Scheduled Jobs, add a time zone dropdown to know what time we are entering. It is so confusing!
1 vote -
1 vote
- Don't see your idea?