Last daily update:
2019-08-20 17:06:51 UTC
Last incremental update:
2019-08-20 17:07:38 UTC
2019-08-20 18:15:39 UTC
Hi everyone, we're upgrading the server to a new machine on Dec 19. Expect intermittent outages which could last as long as 24 hours. No action is needed from users during the upgrade. More workunits will be on the way shortly thereafter. Thanks!
Over the next couple of weeks, we will be moving the Cosmology@Home server to a newer machine. No action is needed from you, and we hope to do so with minimal down time. This will solve lots of the work generation and validation problems the site has had over the last several months. It will also set us up to debut some new science applications which we have been working on and thinking about.
For now, we are letting the work queue get emptied out. Once that is done (it already is for camb_legacy), we will move forward and keep you updated.
Thanks for your continued support!
Over the last week we suffered a database corruption due to some disk errors. I've spent the last several days recovering the database from backups and from the corrupted files. Unfortunately, records of workunits from the last several weeks were lost, which means you will not receive credit for any of these jobs. I greatly apologize for this, and we've taken steps to make sure this doesn't happen again. The good news is that this was the only thing which could not be recovered, everything else is fine.
We're continuing to monitor things as the server comes back online, please report any problems you may find here.
The problem affecting the camb_legacy tasks has been resolved. The server also recently received some under-the-hood upgrades which should improve performance and security.
Hi everyone, there's something wrong with the camb_legacy tasks, possibly related to a recent upgrade. I've temporarily disabled them (no validation, no new workunits created) until I can figure out the problem. The camb_boinc2docker application is still working normally.
Hi folks, I've made a few tweaks to the website theme. Hopefully this makes it a bit easier on the eye. These changes should also improve usability on mobile somewhat, although some work remains to be done. For now, please report any theme / user interface bugs you may find on this thread! Thanks!
The power is back and so are we. Happy crunching!
Dear all, due to a scheduled power outage at our hosting site, Cosmology@Home will be offline starting later today (Dec 13) and may remain down for as long as 48hrs. We apologize for any inconvenience.
Hi everyone, graduate student Lo Lee and Professor Peter Darch of the University of Illinois (where we got our start back in 2007) are conducting a survey of Cosmology@Home users for their research. If you have a moment, we'd appreciate if you could fill in the survey at https://illinois.edu/sb/sec/2695704. Your (completely anonymous and confidential) responses may also help us guide the design of Cosmology@Home in the future. Many thanks!
Hi all. As you might have noticed, cosmologyathome.org was unreachable for the last ~2.5 days. This was due to a failure properly reregistering our domain name after moving the server from one university to another earlier this year. In addition, as we couldn't contact you via the website, we tried to send out email notification of the issue, but due to a problem with the server software many of you did not receive this email. We'll work on addressing this so in the future it doesn't happen again. All should be back to normal now, although it may take as long as another 24 hours for the domain to fully propagate everywhere. We greatly apologize for this extended down time.
Hi folks. Due to a mix-up related to having moved servers from one university to another earlier this year, we failed to properly renew the cosmologyathome.org domain, which expired today. We are in the process of seeing what can be done to fix this. Until then, you can access the website at the temporary URL you are currently at, https://cosmos.iap.fr, and I will post the latest news on the situation here. We apologize for this downtime.
The BOINC Pentathlon is complete, as is our own Planck contest! You can see the results of the Planck contest here and of the Pentathlon here. Congratulations to users Mumps[MM], UofS-Computer-Science, and Bryan, and to team SETI.USA (who also won the Pentathlon)! You guys will be added to the list of thank-yous in our paper.
In the end it was a quite exciting race for the 3rd place user spot, Bryan managed to overtake KaptainBlaZzed with only 45 minutes left in the contest last night! Well done to both of you and many thanks to all of the thousands of users and hundreds of teams that contributed. This was far and away the most computing power I've seen at C@H since I've been managing it, and the workunits you computed will greatly help us improve our work. Stay tuned for a future post about these results.
Hi all, in an effort to keep the server functioning smoothly (its operating pretty close to capacity with all the extra traffic from the Pentathlon) I've disabled creation of new camb_boinc2docker jobs until the contest is over on May 19th. All your in-progress camb_boinc2docker work will finish and be validated, this only means no *new* work for this application will be sent out by the server. If you wish to continue crunching Virtualbox applications (like camb_boinc2docker was), there should be enough planck_param_sims jobs, just make sure you haven't explicitly disabled this application in your preferences. Good luck and happy crunching!
The in-progress standings of the Planck contest can now be seen here! (note: those standings include only Planck jobs, the Pentathlon counts any Cosmology@Home job).
Additionally, validation had been lagging a bit, but that is now fixed. It should take about 48 hours for all the jobs that had piled up waiting on validation to finish and for credit to be granted. There is also a shortage of work for the camb_boinc2docker and planck_param_sims applications, which I'm currently working on now and I expect to be fixed by tonight. Apologies for this!
And in addition to the new Planck jobs, C@H is the "Marathon" project for the yearly BOINC Pentathlon contest! You can find out more about the Pentathlon here.
To slightly sweeten the deal, we'd like to add a few more thank-yous to the paper, so we will again take the top 3 users and top team (excluding the winners from last time) and add your names to the paper. Although you're welcome to start crunching Planck jobs now, the contest will be only considering Planck jobs returned during the Pentathlon which runs from May 5th to May 19th (see above link for a countdown). We'll post an updated page to keep track of the leaders this week. Good luck everyone!
Hi folks. Last year, you guys helped analyze results from the Planck satellite by having your computers comb through millions of simulations of our universe, to help us understand if some curiosities in the Planck data were just statistical flukes, or if they were hints of a new discovery. In a bit more detail, the science behind these jobs was described here, and your results allowed us to write a paper which explored these questions (which we'll summarize in a future post). Along the way we also had a contest for who contributed the most computing power to these jobs, and we ultimately thanked winners MaDdCoW, 25000ghz[Lombardia], Rally1965 and team BOINC.Italy in the paper.
For those that aren't familiar with the way publication works, first a paper is submitted, then a referee replies with comments and requests for revisions, then the authors respond and the paper is eventually accepted or rejected. This is the stage we're currently in now, working on this response. As part of the response, it would be useful to analyze some more simulations, so starting today and continuing at least for a few weeks, we'll be adding back a bunch more Planck jobs.
Thanks again to all who contribute and happy crunching!
Hi all, apologies for the extended downtime we had the past few days. We had a hard disk failure that, since it came over the weekend, took a bit longer to resolve than usual. We are back now and there was no data loss, however its possible some of your clients missed reporting deadlines (since we were unavailable). If you feel you lost significant credits because of this please don't hesitate to contact me. We're working on making sure similar errors in the future can be resolved quicker. For now, sorry again, and happy crunching!
And we're back! The server move to France appears to have gone well. We'll continue monitoring things closely to make sure the new machine is working well. Please let us know if you find any problems, and happy crunching!
Tomorrow the C@H server is packing its bags and getting ready for a big move! After 10 great years of being hosted at the University of Illinois in Champaign-Urbana, we are moving overseas to Paris on a server at the Institut d'Astrophysique de Paris, where both myself and Ben are currently working. You can read more about this in an updated welcome letter from Ben.
At 2pm UTC on Wed Dec 14, we will take the server offline. The process may take up to 24hrs, during which your clients will still process pre-downloaded jobs, but will not be able to communicate with server. Upon finishing the move, you clients will reconnect automatically, and no actions from you are necessary. For the ultra-curious, I will try and post some live updates tomorrow on my twitter.
Hi guys, over the last several days the site has been suffering from what ultimately was a database error that caused several of the problems some of you have been posting / emailing about. Sorry for the slow response with the holidays here, but we should be fully back online at this point. Until everything is flushed out, there may be some workunits which will be screwed up as a result unfortunately. If you feel like you lost a significant amount of credit as a result of this, or if you notice anything still not working, please don't hesitate to PM me with more details. Happy crunching!
We have added the option to access the C@H website more securely via "https", if you'd like to do so, you can update your bookmarks to use https://www.cosmologyathome.org. For now this remains purely optional (for example, if you don't know what https is, you can ignore this). At some point in the future, we will likely move to requiring this, but we'll give warning and post more details then.
Also, I've fixed a number of minor cosmic things about the site, most notably 1) the BBCode buttons when posting to the forums now work 2) a number of pages which were rendered too wide should now correctly fit the screen. If, as you're using the site, you find any other glitches, please don't hesitate to take a screenshot and post on the forums so we can fix things! Thanks all! (Edit: ha, oops meant to say *cosmetic* things above, although I'll leave in the somewhat appropriate typo)
In February we started running a new application. Today the paper making use of the results that thousands of you guys calculated is out! Look here in the coming weeks for more posts detailing exactly what we found. Until then, you can see the paper here. On behalf of the C@H team and of the Planck collaboration, thanks again everyone!
The two Docker-based apps have just been upgraded to 2.01, notably with 1) support for Virtualbox 5.1 and 2) Fixed the "Unknown option: --lpt1" on MAC causing some jobs to fail.
The camb_boinc2docker and planck_param_sims apps are getting an upgrade which should reduce even further the number of failed jobs. Read more about the upgrade in the comments.
A few updates on the status of Planck jobs and some pretty graphs. Read within!
Thanks everyone for participating in the Planck contest! Today we picked the winners, big thanks to the top users MaDdCow, 25000ghz, and Rally1965, and the top team BOINC.Italy, who we will mention in our paper. You, along with the ~2500 other users and ~300 other teams contributed a ton of computing power to allow us to do this analysis over the last several months.
The contest to be thanked in our upcoming Planck paper will end on May 5 at 13:00 UTC. Good luck to everyone in the final two weeks!
Thanks to you guys we've almost hit the 10,000 Planck parameter simulations we wanted to do. Read within what we've learned from these, where were going next, and about a few minor updates to C@H.
I'm working through figuring out why some uprgades didn't quite go as planned, the server may be a bit degraded the next few hours, including download errors and lack of new tasks. Hopefully should be sorted out soon.
The new app is live! Read more about it in this post. The top three users and top team will be thanked in our paper, track stats here. Come help us figure out something new about our universe!
For anyone interested in the inner workings of C@H, you might be interested in this 10 minute screen cast I just posted. Its pretty technical, but shows what running a server and running jobs kind of looks like from our end. And of course if anyone with technical expertise wants to contribute to C@H with your code, not just your CPU cycles, we welcome that too! (just get in contact with me)
Today at 10:30 EST there is a press conference from the LIGO collaboration. The rumors are that they will announce they have detected for the first time ever gravitational waves, in this case from the merger of two black holes! There are also rumors they will announce nothing because their policy is to insert fake signals into their data and not tell those doing the analysis until the very end (a so called "blind analysis"). Tune in to find out, this should be an exciting event! Check out here and here for two nice descriptions of gravitational waves of varying levels of technicality.
Today we are releasing a new app (in beta for the moment) to help analyze Planck data. Read more about it in the comments. Start crunching this app now for a chance to receive thanks in the acknowledgements of our forthcoming paper!
As of right now the camb_boinc2docker application is no longer in beta, meaning everyone will receive these jobs by default unless they have specifically chosen not to. Please read the FAQ and don't hesitate to post on the forums if you run into issues. Thanks everyone and happy holidays!
Here's the video and slides of my talk at the Dockercon conference in November giving some details about C@H. There's a little bit on the science side and a little bit on the technical side for those of you that are interested!
We were down for the last ~24hrs due to some networking issues which are now fixed. We apologize for the inconvenience.
On Dec 3 I will remove the beta status for the new app ("camb_boinc2docker") which means everyone who meets the system requirements will get the new jobs by default (currently only those of you who have selected "Run test applications" in your C@H preferences are getting them).
To ensure a smooth upgrade, please take a look at the requirements and FAQ. The most important things are having Virtualbox installed and enabling VT-x/AMD-v extensions. Thanks everyone!
I just attended Dockercon Europe 2015, a huge tech conference for Docker, which is the software that powers the new camb_boinc2docker jobs. I also gave a talk about Cosmology@Home (which I will post a video of as soon as its available). But little did I know as I was listening to the keynote presentation along with a few thousand other attendees that Docker CEO Ben Golub would pick our project to showcase the types of things you can build with Docker. Needless to say I was floored and hugely honored! It was really cool, and thanks again to you guys for volunteering your resources to make Cosmology@Home possible!
Here's video of what he says (link should point to 23:12 right where he starts talking about it). I think he really nicely summed up what is exciting about C@H, and his description of the science we do was impressively spot on!
Upgrade completed successfully.
C@H will be down for upgrade Wed Oct 21 at 12pm UTC. Read more here.
Public beta testing for the new C@H is now open! Read more here.
A new post containing statistics about Cosmology@Home users. Discussion here.
Kevin implemented further fixes to the web site and the server software. The participant profiles can now be browsed through the web site. Spam profiles have been removed. Upload errors should have been cured. Please test and let us know what you find.
A major update has now been completed - the HTML/PHP for the website has now been updated to the latest version. This was long overdue and has hopefully closed some security loopholes. We are now cleaning out a backlog of spam user profiles, etc. Please let us know if anything is not working as expected as a consequence of these updates..
As part of the recent updates (more storage, more recent kernel) we finally fixed a longstanding problem: the community preferences are now working again! Check out our main post on the message board for the project announcements.
We updated the featured concept to Dark Matter, with a cool image of the bullet cluster. At the bottom of the featured concept on the home page there is a link that will take you to a thread on the message board where you can comment and ask questions. Enjoy!
After installation of 4TB of disk space and a re-organization of the file system, and a few updates and bug fixes, such as updating the data base to fix an annoying and persistent problem with setting user preferences, we are back in production. Thank you for your patience and keep crunching!
We are running again, with a temporary fix for the disk space problem. 4 TB of disk space have been ordered. The team data base has been repaired. We will keep you updated when the filesystem upgrade happens, which we will take as an opportunity for a few additional updates.
As some of you have pointed out, we are currently experiencing problems with the file system which are affecting the module that assimilates the results of your computations. We are working to resolve these problems and will endeavor to keep you updated as we progress. Please bear with us as we try to resolve this problem.
Please join Ben in welcoming Faik, a new member of the cosmology@home team.
Congratulations to Saul Perlmutter, Brian Schmidt and Adam Riess! Read Ben's post about today's Physics Nobel Prize.
In honor of today's Physics Nobel Prize winners our Featured Concept on the home page now discusses type Ia supernovae, exploding white dwarfs that were key to the discovery of the accelerated expansion of the Universe.
We invited the other members of the US Planck team to participate in the user community of cosmology@home. Please fill in this survey to comment on what features you would like to see most.
Updated our Featured Concept. This was long overdue and will now occur semi-regularly. The more popular demand, the more frequent the updates. Let me know by posting on the message boards what concepts you would like to see discussed.
Fixed several bugs on the site caused by migration from *.uiuc.edu to *.illinois.edu, plus a few others. Watch for further updates.
Planck completes its baseline survey - now in extended mission. ESA announced mission extension until 2014! Follow the Planck mission at this website.
A total of 5000 credits have been added to profiles of all active Cosmo@home users. This is meant for expired/failed WUs due to project down-time and to compensate for the earlier low credit per WU payment for resource intensive CAMB 2.16 WUs. Also, the time bound for these new WUs has been increased from 10 days to 15 days which should be enough to complete crunching them even on machines ~5 years old.
We have restored CAMB 2.16 apps successfully for all platforms so all project specific errors (Signature and/or Checksum) should no longer be present. Also, most WUs for camb 2.16 are more resource heavy than the previous WU's for 2.15 so we have increased the credit for each WU from 140 to 200. If there are any disputes about this we will be happy to address them. If you feel that the credit being given for the 2.16 app, WUs is too little/ too much please start a thread related to this on the Wish List message board. Thank you.
After replacing a damaged RAID drive, the server kept throwing I/O errors and kernel panics, which required us to have the motherboard and RAID controller replaced as well. Hopefully this last round of hardware replacements will solve the problem and we will be back up today.
The website is mostly up and running at the moment. We'll have the project daemons up as soon as possible.
The Cosmology@home server has been suffering abnormal crashes ever few hours since the campus power outage on March 8, 2009. We will be taking to server offline soon for maintanence and repairs during which time Cosmolosy@home will not run. This website and the forums will also not be accessible at that time. We are working as fast as possible to get the project back up and running to its full potential.
There had been a power outage at the University of Illinois campus today for approximately 12 hours due to strong winds knocking down large power lines in the county. Due to this some of you may have experienced download problems from Cosmo@home over the course of the day but the power has now been restored and everything should be back to normal. If someone still experiences problems please let us know through the forums. Thank you.
A problem which was causing the test parameter file generator to output empty parameter files has been corrected. I just tested a new test WU on my machine and it seems to be working file. I'll be creating another 500 WUs to test the app on other architectures. The credit for these WUs (excluding the 5 that I just made to test whether parameter file generation was working correctly) will be 3 times the normal amount.
We're in the process of debugging the test app. There will periodically be test WUs released in small batches to test the new app version. The new WUs should require lower accuracy and should complete more quickly than the last batch.
Credit will be given out for successfully completed WUs (that is, WUs without errors). However, we'd like to remind everybody that credit is not guaranteed for test WUs. Test WUs help us to release new versions of CAMB and keep C@H going; be aware, however, that these WUs can (and often do) fail, so if you would rather crunch regular WUs and not have to deal with the hassle of failed WUs, then disable test work in your project preferences. Crunching test WUs will never be compulsory. That being said, however, we will make every effort to reward users who donate their CPU time to help us test.
A new test application has been released and 500 test WUs created. Please enable test work in your project preferences if you would like to help us test this new version so we can put it into production.