Crash "Apps may not schedule more than 100 distinct jobs"

[kvillnv]

Hi, I am testing an app since a few days I have been scheduling tasks with workmanager everyday. today I added a new one and I got a crash with the following log (all my tasks of the previous days were already finished).
Now the app can not start anymore, unless I delete all the app's data.

1-Does this mean I had 100 tasks during the last few days?
2-Or does this mean my last task was accidentally launched >100 times?
if 1, why are we limited to 100 tasks

Is it a known issue? any help would be appreciated. thanks

06-01 13:11:09.180 696-4700/system_process W/JobSchedulerService: Too many jobs for uid 10095
E/AndroidRuntime: FATAL EXCEPTION: pool-2-thread-1
Process: tv.ebox.freetv.debug, PID: 26258
java.lang.IllegalStateException: Apps may not schedule more than 100 distinct jobs
at android.os.Parcel.readException(Parcel.java:1691)
at android.os.Parcel.readException(Parcel.java:1636)
at android.app.job.IJobScheduler$Stub$Proxy.schedule(IJobScheduler.java:158)
at android.app.JobSchedulerImpl.schedule(JobSchedulerImpl.java:42)
at androidx.work.impl.background.systemjob.SystemJobScheduler.scheduleInternal(SystemJobScheduler.java:85)
at androidx.work.impl.background.systemjob.SystemJobScheduler.schedule(SystemJobScheduler.java:64)
at androidx.work.impl.Schedulers.scheduleInternal(Schedulers.java:98)
at androidx.work.impl.Schedulers.schedule(Schedulers.java:69)
at androidx.work.impl.utils.EnqueueRunnable.scheduleWorkInBackground(EnqueueRunnable.java:109)
at androidx.work.impl.utils.EnqueueRunnable.run(EnqueueRunnable.java:81)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1133)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:607)
at java.lang.Thread.run(Thread.java:761)

[biliboss]

@kvillnv i had this issue in my tests, then i started playing with uniqueWork and the problemas was gone.

[kvillnv]

@biliboss thank you for your fast response.
I am going to add beginUniqueWork() before .enqueue()

[kvillnv]

@biliboss it seems it did not fix my issue.
here is how I use uniqueWork

val work = OneTimeWorkRequestBuilder()
.setInputData(data)
.addTag(mediaName)
.setConstraints(constraints)
.setInitialDelay(initialDelay, TimeUnit.SECONDS)
.build()

WorkManager.getInstance()
.beginUniqueWork(
mediaName,
ExistingWorkPolicy.REPLACE,
work).enqueue()

I tried unique tags with ExistingWorkPolicy.REPLACE
and I also tried with same tag and ExistingWorkPolicy.APPEND

UPDATE:
first task I create: one task is created
2nd task: 2 tasks are created
3rd task, 3 tasks are created..... and it continues to grow

But my method WorkManager.getInstance() .beginUniqueWork(...).enqueue() is called only once

[biliboss]

@kvillnv could you confirm that problem occurs only after open your app 100 times?

In your code, mediaName is a Constant?

Are you sure that your Worker isn't stay in a infinite loop?

[kvillnv]

@biliboss
-The problem can occur on a fresh install: Actually everytime I create one task, it seems that all the existing tasks are also re-created so when I have 8 tasks, when I create the 9th, the 8 others are also recreated, so I end up with 17 tasks. I can see that because there is a log from Android's SystemJobScheduler class that is repeating as many times ("Scheduling work ID %s Job ID %s"). I am able to recognize the WorkIds of all the previous tasks, and I can see the JobID incrementing (to 17 in this case).
-When I used ExistingWorkPolicy.APPEND, mediaName was a constant, and when I tried ExistingWorkPolicy.REPLACE, mediaName was a variable with a different value everytime.

I confirm,there is no infinite loop, according to my logs, my method WorkManager.getInstance() .beginUniqueWork(...).enqueue() is called only once.
But I can see the method scheduleInternal() from Android's SystemJobScheduler is repeating as many times as the task is created.

What ExistingWorkPolicy TAG fo you use in the beginUniqueWork() ?
Do you use the method the same way I wrote in my previous message?

[biliboss]

Hello @kvillnv.

I used KEEP in the start of program, and APPEND if it's called by itself (worker).

Do you already confirm if your Worker is returning a result? We saw this log here.

06-04 10:05:53.398 25529-25910/? D/WorkerWrapper: Worker result FAILURE for 417f62a3-0f24-4fb9-8884-cb440adae758

[kvillnv]

Hi @biliboss
The problem occurs before the result (the tasks are scheduled later in the day).
If I wait for my first task to be executed, when I create the second one I won't have the repeating issue.
And when I have the duplicated tasks bug, it seems that each task will be executed only once.
When my tasks are executed I always have result SUCCESS
Thanks for your help.
My mistake must be somewhere else. I'm going to recheck again all my implementation.
I thought there was a bug in the alpha version.

[kvillnv]

This is a bug.
When scheduling a new job in the future and already having several other job scheduled : the whole job list is duplicated every time.

[saulmaos]

i confirm that bug, i have the same issue, my already canceled tasks are being printing in Logcat just like yours

[kvillnv]

I decided to remove workmanager from my project.
Finally I implemented Android JobService and I schedule with JobScheduler and JobInfo.
Now everything works fine.

Maybe workmanager is good for periodic tasks or short-term scheduled tasks in chronological order. But when you want to schedule at specific times differents jobs before or after another one it does not work well.

[biliboss]

I rolled back to the Alarm Manager after figured that workmanager isn't have a guaranteed timed execution, and so a little buggy yet.

I think that Work Manger isn't production ready.

[saulmaos]

@kvillnv i used it for periodic tasks and got the the bug mentioned before.

[maiconhellmann]

I got the same bug.

[matemink]

Got the same bug for periodic tasks