Please, I have many work units in the Boinc Manager in “Ready to report” and “Computation error”.
This has been going on for several days and when I click on “Update”, the work units do not go to the server.
What should I do ?
Should I wait ?
Deadline for work units: January 1, 2025.
Thanks !

I too have been getting lost of tasks finishing with a "Computation error". SOME tasks do complete successfully, but a lot haven't.

Also, I've noticed that the tasks that don't complete properly have been crunching for nearly 4 hours then fail with the computation error. The successful tasks complete in just over 4 hours.

The tasks are still stuck in my Boinc Manager and they don't go back to the server, even successfully calculated tasks.
The deadline was January 1st !
I don't know why everything is blocked ?
I can't do any more calculations for this project !

The time it takes me to calculate a task is 2 hours for those that succeed and 39 seconds for those that go wrong.

Do you still get Computation errors on new tasks? If so, can you please try restting the project?

Unfortunately, I don't receive any new tasks !
And if I restart the project, I risk losing 857 “Ready to report” tasks.

On the server, your tasks are in status "No reply". I don't know why they are not being reported.

Can you check the logs of your BOINC client/manager for comunication or other errors?

Is this what you need ?

sam. 04 janv. 2025 11:44:06 | LODA | update requested by user
sam. 04 janv. 2025 11:44:09 | LODA | Sending scheduler request: Requested by user.
sam. 04 janv. 2025 11:44:09 | LODA | Reporting 2007 completed tasks
sam. 04 janv. 2025 11:44:09 | LODA | Requesting new tasks for CPU and NVIDIA GPU
sam. 04 janv. 2025 11:44:14 | LODA | Scheduler request failed: HTTP internal server error

I extended the deadline and reset the status of your workunits. Can you please restart BOINC and update the project?

I restarted the computer and updated the project.
Sorry, the answer is the same !

Should I reset the project ?
So much for my 857 good tasks !

sam. 04 janv. 2025 13:22:32 | LODA | update requested by user
sam. 04 janv. 2025 13:22:37 | LODA | Sending scheduler request: Requested by user.
sam. 04 janv. 2025 13:22:37 | LODA | Reporting 2007 completed tasks
sam. 04 janv. 2025 13:22:37 | LODA | Not requesting tasks: "no new tasks" requested via Manager
sam. 04 janv. 2025 13:22:41 | LODA | Scheduler request failed: HTTP internal server error
sam. 04 janv. 2025 13:22:52 | LODA | work fetch resumed by user
sam. 04 janv. 2025 13:23:03 | LODA | update requested by user
sam. 04 janv. 2025 13:23:09 | LODA | Sending scheduler request: Requested by user.
sam. 04 janv. 2025 13:23:09 | LODA | Reporting 2007 completed tasks
sam. 04 janv. 2025 13:23:09 | LODA | Requesting new tasks for CPU and NVIDIA GPU
sam. 04 janv. 2025 13:23:13 | LODA | Scheduler request failed: HTTP internal server error

You will most likly lose your credits when you reset the project. I have extended the deadline to 15 January.
The last contact of your host to the server was reported on 26 Dec 2024, 12:14:06 UTC.

It doesn't help for the existing tasks, but as a general hint, you can set a maximum number of tasks on th website at Project -> Preferences.

Thank you very much Christian for the explanations and the extra time.
I will try again to save these tasks.
And maybe my friends at Alliance francophone will have an idea !

I checked some more metrics. There were no programs found by your computer since 27 Dec. You can see it in this dashboard (select your name):
http://dashboard.loda-lang.org/grafana/?orgId=1&refresh=1m&from=now-15d&to=now
This indicates that your computer didn't execute tasks since then (but I'm not 100% sure).

OK, many thanks Christian.
I'll wait until the new dead line and see if the tasks go to the server.
If they don't by then, I'll have to detach then reattach the project and lose them !

Many projects cannot handle a user reporting that many tasks at one time. That may be what is happening here. Try reporting just (say) 100 tasks at a time. You do this by adding this to your /var/lib/boinc/cc_config.file:

<cc_config> 
	<options> 
		<max_tasks_reported>100</max_tasks_reported>
	</options> 
</cc_config> 

Be sure to tell BOINC to re-read config files after you add it. Or quit/re-start BOINC. Either way should work.

Bingo !
It works.
Many thanks zombie67.
The work units were sent to the server.
But none have been validated !
I guess all those calculations were for nothing !
That's okay, I really like this project and I'll get back to work for Loda soon enough !