The rosetta@home is out of work units on all platforms. Can we expect another project?
Hi Adam. The Rosetta status page shows all services online and work available; I wonder if the issue you observed has been resolved? -- Do you still see a shortage on your end?
(Is your system Windows, Mac, or Linux?)
Guys, I'm seeing exactly the same problem .. on Linux machines located in Australia and the US.
The servers are working, but there is no WU.
See the Tasks by application section down the page.
Ok. We'll add another science project for Linux devices. (For Windows, there should be plenty of other work.)
I can't get new Wu on Windows too.
They sent out a couple hundred WUs, but they run out again.
For Windows, though, you should be getting work for other projects though; notably from CE projects.
I can't see another science project for Linux yet; when should we see it?
And .. no more Rosetta tasks coming .. again!
Not sure what's up with Rosetta. We'll activate one or more new proejcts by early next week. (Work here slows down a bit during the holidays; thanks for your patience.)
We've added ClimatePrediction to hosts running Rosetta, so this should keep you busy with work at all times :)
Thanks for your participation and happy new year --
How do you decide which project to add?
The link missing a colon.
The ClimatePrediction gives too big WUs. It shows 44 days to finish, but the deadline is only 13 days away.
The ClimatePrediction thing doesn't give me Any tasks! And Rosetta has also stopped allocating tasks. Matt, can you please revisit this whole issue.
Same issue with the ClimatePrediction on Linux. No WU.
Plus BOINC shows this message: This project is using an old URL. When convenient, remove the project, then add https://climateprediction.net/
Yeah, seeing the same here. How do we safely remove the project so that it will be automatically added back correctly?
For Linux users we've made the following changes:
Matt, thanks for the update.
World Community Grid, not an option? They have a wide variety of WUs for different platforms.
Still can't see Universe@home or Cosmology@home. And ClimatePrediction has never found any tasks, still set to "Won't get new tasks", and won't disappear.
And Rosetta has stopped getting tasks .. Again :(
Graham, I missed a step in activating those projects for you, apologies. You should see them show up now (may need to manually sync with Charity Engine, if it hasn't happened automatically). ClimatePrediction we will leave there for a bit in case other users do have active tasks. It's a burden on the projects if we detach quickly and don't let jobs complete, so that one will just disappear quietly at some point...
Thanks Tristan. I've just done three successive manual syncs on Linux machines in Australia and the US. No new tasks for Rosetta, and no new projects.
Can you please take another look at this.
Wow, activation still didn't stick. Sorry about that again... it really should work this time! (Confirmed on my own Linux machine, at least).
Yes! I now have both Universe and Cosmology tasks running. Thanks. :)
Another issue. Cosmology seems to think it's running on a VM; it's actually on bare-metal Ubuntu 18.04. Thus:
Fri 04 Jan 2019 18:23:04 AEDT | Cosmology@Home | Task camb_boinc2docker_20017_1546543807.599031_0 postponed for 86400 seconds: Communication with VM Hypervisor failed.
Cosmology runs some work that requires VirtualBox to be installed. That would be surprising if you received such a task when you don't have VirtualBox... but maybe that is the issue?
If you don't have it, getting it is as easy as:
$ sudo apt install virtualbox
I already have virtualbox. Presumably I need to start it at boot time, running in non-gui mode as user 'boinc'. How should I do that?
Ah, think I've figured it. I've just installed boinc-virtualbox; watch this space :)
Well it seemed like a good idea. But it still didn't work thus:
Sun 06 Jan 2019 20:27:56 AEDT | Cosmology@Home | Task camb_boinc2docker_6316_1546765802.994319_0 postponed for 86400 seconds: Communication with VM Hypervisor failed.
Can we perhaps tell Cosmology that we don't want to accept Docker/VM jobs on this machine?
Note that Cosmology is working fine on another of my machines .. a basic CentOS 6.8 machine without Virtualbox.
It's possibly a file permissions issue... there are reports of this with other Debian derivatives. Maybe check permissions under /var/lib/boinc-client, particularly if there is a hidden .config directory. That may be where VirtualBox is trying to store data for the boinc user, which may not work if owner/group permissions are set to root instead of boinc.
(This may explain why it would work on some systems but not on others.)
Think I've solved it. An old Boinc installation on my machine was done under /opt/BOINC. That directory was subsequently removed, but its location was retained in /etc/passwd as the Boinc home directory. Boinc was smart enough to figure that it should use /var/lib/boinc .. but boinc-virtualbox wasn't.
Solution .. fix /etc/passwd then remove and reinstall boinc-virtualbox.
Cosmology@Home | Requesting new tasks for CPUCosmology@Home | Scheduler .. completed: got 1 new tasksCosmology@Home | Started download of params_de201167 ..Cosmology@Home | Started download of vbox_job_8c11dc ..Cosmology@Home | Finished download of vbox_job_8c11d ..Cosmology@Home | Started download of boinc_app_0f8af ..Cosmology@Home | Finished download of boinc_app_0f8a ..Cosmology@Home | Finished download of params_de20116 ..
Why are you going to remove the Climate Prediction project? It seems it has a more important cause then Cosmology.
BTW, when are you going to remove the project? It has huge tasks, which is not a problem because they have a 1-year deadline.
Good to hear, Graham, and nice work tracing down the root of it!
Adam, we are removing CPDN based on your earlier feedback. :) Deadlines were too short and no work was available for Linux.
As for Cosmology, if we are able to pry back the mysteries of something like "dark matter," we may find a much larger world of understanding that benefits the topics of climate and cancer as well. As Alexander Fleming said of his discovery of penicillin, "One sometimes finds what one is not looking for."
Thanks. Sorry, I misinformed you earlier about the deadline, because I misread the deadline. I didn't notice the year, so actually, the deadline for the WUs are 1 year :)
running windows and have recieved universe@home work even though in earlier post it was stated it was for linux but now i have a bunch of completed universe@home tasks that are ready to report but wont report and clear them keeps saying comm deffered for long periods of time then when time runs out it says scheduler request pending then goes back to comm defered for twice as long as previous have had completed work for 2 or 3 days now that will not report and vanish
I have a lot of finished task from universe@home too. It seems their server is down.
It's ok now. There was maintenance for 3 days.
One of my computers needs at least 14-18 more days to finish all of the Climate Prediction tasks. Can you guys keep the project for that long?
Sure, that shouldn't be a problem. I'd guess you're not alone in still having active tasks there.
Every time any of my machines does a "Synchronize with Charity Engine" it sets status for Universe@Home to: "Won't get new tasks". So what's happening here?
There is an unbalanced amount of credit coming from that project, so it has been suspended for a bit while we look into how to handle it.
I'm not getting any more VM Cosmology tasks on my Linux machine(s). Has something changed in the last couple of weeks?
And did we ever get Universe@Home going again?
Nothing has changed with project assignments, but I'll look a bit more into Cosmology. Currently their site is down, but there doesn't seem to be an issue with work going out in general.
Universe@home is still suspended. We'd have to change the way points are calculated to accomodate the wide variation in credit being awarded (so maybe a future improvement).
I've got VM Cosmology tasks again, so that issue seems to have resolved itself. But I have noticed that the Cosmology site is often down, sometimes for twelve hours at a time.
And now it's more than 24 hours, which means that one of my low-power Linux machines has been unable to get any work for most of that time.
Maybe they'll have it back up in another day or two? (They probably know about the outage already, but I've sent a note to be sure).
Cosmology seems to be incommunicado .. again :(
Seems to be back up, another of their regular downtimes?
Cosmology is down again this weekend. One of my machines will only support Cosmology jobs, so it's effectively idle again.
Was there any response from the note sent to the Project managers?
We've broadened attachment to LHC, which should run on any system with VirtualBox installed. Granted, that won't work for everyone either, but hopefully helps a bit!