I hope this reminds the people who are running the project to load new tasks to the server thanks.

At time of posting Workunits waiting for validation 159322 I am not 100% certain however I believe the server is working through the validation backlog when this has lowered new work will be created. I have no idea what the timeframe will be. It may take it just over 6 hours as this is how long it says the backlog is however this may not be the case

I don't know why the backlog is not clearing. I've stopped new task creation for now to see what happens over the night.

Appreciate the information. Hopefully it will start to clear overnight. In the last 30 minutes 16 results of mine have changed from pending to validated

I don't know why the backlog is not clearing. I've stopped new task creation for now to see what happens over the night.

Sergei, there are now (21:55 GMT) over 3k new tasks available. Were these created before you stopped new task creation?

Yes, they were created before. It's a huge backlog overall.

A large number of tasks to be approved
Validation pending 24 975

---

Workunits waiting for validation 155015

Still bad

Looks like the servers have no new WU's. Time to move on to a new project.

All WUs are completed now, there will be no new tasks. There are still some pending "in progress" tasks that prevent the progress from reaching 100%, but they will be cleared as "timed out" or "cancelled" soon.

Does this mean that we have successfully completed this project? If so, good job everyone.

Almost, all tasks have been sent out it seems. Just waiting on everyone to finish them and return them.

This could take a while if people have let some of their tasks time out. With such long deadlines we could be waiting a week or more for the task to timeout and get resent.

Time will tell, but so far with less than 3k tasks in progress it's been dropping really fast. Had almost 30k tasks in progress 12 hours ago.

Some people have been bunkering tasks like crazy for the past few days.

I didn't want to let it drag on for several weeks, so I kept issuing extra tasks for remaining WUs until they all got validated. Then I just force timed out remaining tasks. It's all clear now.

Almost, all tasks have been sent out it seems. Just waiting on everyone to finish them and return them.

No, any tasks returned now are already worthless from the project's perspective, as all workunits have been sent out to lots of additional hosts and already met their quorum, and not even validated any more, as their deadline was moved up and they are too late now. It might make sense for the project to set send_result_abort and actually cancel the remaining tasks, as otherwise tasks that are already running on the clients will run to completion despite being superfluous, even if they happen to contact the server now. The updated deadline is not propagated to the clients, so this is not sufficient to cause them to abort any task in progress.

There are less wasteful ways to wrap up a project, but at least this one was actually finished unlike far too many others that just were abandoned or disappeared without any notice.

I absolutely agree with pschoefer. This chaos caused by the zigzag course in the past few days was more than unnecessary from the cruncher's perspective. I personally lost tasks worth millions of credits, most of them already finished, and they all got invalidated. Furthermore, this also prevented me from reaching my next milestone. I'm a bit disappointed now.

That's the risk one takes when modifying the client to get more task than the project set amount.

I do hope the admin leaves the server running for a little while so stats sites have time to collect the final results through.

That's the risk one takes when modifying the client to get more task than the project set amount.

Actually, you do not even need to modify the client to do what walli (and myself, guilty as charged :)) did. While I still agree with you and am not disappointed myself that I missed my final far-reaching goal, some resources of participants who were just crunching along without becoming creative to reach any particular goal were also wasted: e.g., almost no CPU that got assigned a new task on the last day had any chance of completing it in time. There would have been options to avoid this and still speed up the processing of the final workunits, but well, it's too late now.

The project was set to send max 1 task per CPU and max 4 tasks per GPU. Normally it's not so much wasted resources, unless you start bunkering tasks. Also think that you are disappointed only because you didn't get to crunch tasks you shouldn't have in the first place (other crunchers should've gotten them - they should be the ones disappointed).

The server will be running in the foreseeable future. It hosts the amicable numbers database, so it's not going away just because the project ended.

Yeah, modifying the coproc file to fake number of GPUs is similar to modifying the client.

All this wouldn't have happened had the crunchers not started trying to horde tasks on the last day anyway.

Reaction and opposite reaction sort of thing.

I was hoping to reach 10B, but I didn't. Can't blame anyone but myself for not doing it sooner. I had years to push towards that goal, but chose the last few months.