Exchange partially processed packet?

Author	Message
old_user466193 Send message Joined: 13 Aug 07 Posts: 9 Credit: 17,815 RAC: 0	Message 30723 - Posted: 26 Sep 2007, 23:42:48 UTC Hi, I\'ve been running a climateprdiction app in my Bionic client for a while now and it\'s at 22% still with 1400hr s to go. I\'ll be going overseas for 6 months soon and as i understand it this means that the job number will probably be given to someone else due to my inactivity. What i\'m wondering is whether its possible to pass on the data i\'ve processed to someone else so they can continue it? Its seems a waste to hahve processed all these hours (511) to just have it someone else start from scratch. thanks orfeo ID: 30723 · Reply Quote

mo.v Volunteer moderator Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0	Message 30724 - Posted: 27 Sep 2007, 0:29:33 UTC If you have a friend prepared to download boinc and complete your model, yes it can be done and we can explain how. CPDN doesn\'t organise this sort of thing though. If you haven\'t got a friend able to do this, I suggest the following. Look at who is sharing this workunit with you: http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=6078981 It was sent to 8 computers and you\'re #8 in the list, processing successfully. Computers #1, #2, #4 and #6 have crashed the model (I didn\'t check whether any of them have later restored it from a backup to continue processing it). #3 and #7 haven\'t started it. Computer #5 has reached 2028. This is a typical situation for a workunit - most members with processing problems don\'t come to the forum to get help, so a lot of models crash. When you return from your trip you could look at this workunit page again to see whether any of the other computers have completed the model (we can help you check this, in this same forum thread). If another computer has completed it, abort it then. If nobody has succeeded, continue the model. I hope you\'re going somewhere nice for six months. Cpdn news ID: 30724 · Reply Quote

MikeMarsUK Volunteer moderator Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0	Message 30731 - Posted: 27 Sep 2007, 7:04:15 UTC IF you wait until the model reaches just over 25% (1960) before aborting it, the model will send a \'restart dump\' to the servers. This can theoretically be used to continue onwards from that point. I'm a volunteer and my views are my own. News and Announcements and FAQ ID: 30731 · Reply Quote

old_user466193 Send message Joined: 13 Aug 07 Posts: 9 Credit: 17,815 RAC: 0	Message 30739 - Posted: 27 Sep 2007, 12:34:23 UTC - in response to Message 30731. IF you wait until the model reaches just over 25% (1960) before aborting it, the model will send a \'restart dump\' to the servers. This can theoretically be used to continue onwards from that point. cool, thanks guys, i should reach 25% so maybe i\'d consider dumping it, but i worry that may actually mean that the work gets lost and wasted so maybe i\'ll just hang onto it for the 6 months and see how other people go. Unfortunatly i haven\'t been able to convince my friends to get on bionic :( ID: 30739 · Reply Quote

mo.v Volunteer moderator Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0	Message 30742 - Posted: 27 Sep 2007, 15:09:21 UTC Every complete 10-year period, which results in a zip file being sent to the server, is used for the research, so what you\'ve done will definitely not be wasted. It\'s not easy to convince everybody that boinc projects are a good idea, is it? I feel the same way about bungee-jumping. Cpdn news ID: 30742 · Reply Quote

mo.v Volunteer moderator Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0	Message 30743 - Posted: 27 Sep 2007, 15:17:58 UTC Hang on a sec - this is an 80-year 2000-2080 model which has reached about 2018 or 2019. So it won\'t reach the 40-year restart dump until 2040 which is over 20 model years away. Before you go away the best thing would be to get it past Dec 2020 or Dec 2030 depending on how much time you have left. Cpdn news ID: 30743 · Reply Quote

MikeMarsUK Volunteer moderator Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0	Message 30746 - Posted: 27 Sep 2007, 20:14:16 UTC \'Doh!\' I keep forgetting that there are other options apart from 160-year coupled models now... I'm a volunteer and my views are my own. News and Announcements and FAQ ID: 30746 · Reply Quote

Les Bayliss Volunteer moderator Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0	Message 30747 - Posted: 27 Sep 2007, 20:33:43 UTC ... but i worry that may actually mean that the work gets lost and wasted ... That particular model may get lost, but so do thousands of others for various reasons. And there are millions of combinations for the researchers to try, most of which never will be. As for wasted, data is returned near the end of each model year, and if you can get just past a 10 year period, a larger amount of data will also be uploaded as a zip file. So there will be little \'waste\' from you stopping before the end. Backups: Here ID: 30747 · Reply Quote

old_user466193 Send message Joined: 13 Aug 07 Posts: 9 Credit: 17,815 RAC: 0	Message 30753 - Posted: 28 Sep 2007, 14:01:57 UTC - in response to Message 30742. Every complete 10-year period, which results in a zip file being sent to the server, is used for the research, so what you\'ve done will definitely not be wasted. It\'s not easy to convince everybody that boinc projects are a good idea, is it? I feel the same way about bungee-jumping. yeah when i get into my uni break i\'m going to email the deal to see if i can convince him to put it on the library computers, 100\'s of computer just sitting there while people look at email and the library catalogue orfeo ID: 30753 · Reply Quote