| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: repetitively resent declared 'unsupported' and discarded |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 15
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
i set up 2 pc at the same time, ubuntu 9.10x64, boinc client version 6.10.17 for x86_64-pc-linux-gnu
1 (AMD) is running great, faah,hpf2,& a hfcc right now, the other (Intel), however; Requesting new tasks Scheduler request completed:got 1 new tasks message from server: resent lost result CMD2.... [error] App version has unsupported platform i686-pc-linux-gnu [error] No application found for task: i686-pc-linux-gnu 614 ; discarding Requesting new tasks Scheduler request completed:got 1 new tasks message from server: resent lost result CMD2.... [error] App version has unsupported platform i686-pc-linux-gnu [error] No application found for task: i686-pc-linux-gnu 614 ; discarding etc. I tried the usual. reset project, attach/detach kinda stuff eh but i am at a loss now as to what to try p.s. those 2 quads i set up i attached to 40 projects and 3 others reacted differently; qmc einstein & climate all said "no work available for your platform to (intel) sent tasks to (AMD) which promptly crashed will your server give up and stop trying to send me the cmd2 (which are not selected anyways) and eventualy send work ? or do i reinstall ubuntu boinc and the other 39 projects under a new hostname to try again here? |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
It's a first I see reported as work being refused for a Linux 64 bit AMD, it loosing the job in transmission and than discarding. At least with the other sciences running successfully we know that the base is correctly configured.
----------------------------------------CMD2 is special as it has to be distributed in very fine grained CPU groups as else the results in a quorum would not validate e.g. broadly, a result of an Intel X would not match the result from AMD Y. Maybe there's a flaw in that table? We'd have to see for a tech's response when coming in tomorrow. If servers eventually realize there's a specific client issue for a specific science I don't know. Given the problem is with only 1 of 5 of your devices, suggest you go into My Grid > Device Manager and create a new group, home, school, work and under custom, deselect Help Cure Muscular Dystrophy 2 [by default all WCG sciences are selected]. Then associate the troubling device with that profile. To see the error from the CMD2 tasks, can you please go to the My Grid > Result Status and click on the error link and post content. ttyl
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Dec 6, 2009 9:28:11 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
:D actualy i have 14 pc connected to wcg, the misbehaving 1 is the only host at location work and work has always been set to accept work from CEP or DDDT only
it sends the file to quickly to catch on the transfer page and stop, and it never shows up on the tasks list. https://secure.worldcommunitygrid.org/ms/devi...s.do?workunitId=113727792 says task in progress |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
okay, that explains part. There is no CEP & DDDT and also FLU phase 1 have finished. Then the "send other work if none of my favs ain't have any" kicks in. You'll have to select 1 or more of the active projects to not receive the HCMD2 anymore.
----------------------------------------That link only gives the quorum header. Maybe you can go to the transfer tab in the BOINC Manager to abort the transmission, causing the task to go away.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Then the "send other work if none of my favs ain't have any" kicks in. is set to don't, work location was for only getting me to gold so machine should get nothing. the download takes 0 time, it doesn't even show on the transfers page. it also never shows up on the tasks page   no biggie, long as its not happening to others. that task is due in a week, hopefully it will give up on my doing it then ;-) or send me the 64bit app to do it with. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
:( I'm getting the same kinda action on another ubuntu 9.10 machine
----------------------------------------an old sempron it's an hfcc task this time cycling away it just failed a hpf2 mz016_00001 "no reply" resent 35 times before its deadline. and thats all its done 'cince the simap ran out (dec 2) what exactly determines the application selected for a task ? is it assigned by your server or the boinc client? hfcc run on 64bit linux so what is telling it to do the task with a application the host wont be allowed to run. it is starting to sound like a boinc client problem [Edit 3 times, last edit by Former Member at Dec 10, 2009 11:45:42 AM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I've asked the techs to look in.
----------------------------------------What goes to a client is determined by: 1. The projects selected in My Projects and the Device Profiles 2. WU availability for the identified project 3. The platform the client is identifying itself with. Now why in this case the 32 bit (the only thing WCG has for all sciences) HFCC runs fine whilst the long known finicky HPF2 does not, in this clients case... that the techs will have to tell. Mind you, read several times 9.10 was not exactly bug free at it's initial release. What BOINC client distro are you running? Synaptic Apps Manager's version?
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
this second problem host has 16k cs so it used to work
it is rejecting the hfcc saying application version has unsupported platform i686... soz i shouilda put that in the first post. i'm running BOINC 6.10.17 downloaded to root , sh'd, and run from the folder it produces down there. all 7 nix boxes are set up this way 1 linux mint64, 5 koala64, 1 koala32 |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Just a try... Reset the projects on both the problem clients [will loose any WCG work in progress]. In the GUI, that's the BOINC Manager advanced view, projects tab, select WCG and hit the button on the left side. This forces the redownload of the science apps, which goes before the work unit downloads.
----------------------------------------edit: Plz check before if in the data_dir sub directory ...\projects\www.worldcommunitygrid.org\ is an app such as wcg_hpf2_rosetta_6.03_i686-pc-linux-gnu wcgrid_hfcc_autodock_6.10_i686-pc-linux-gnu wcg_hcmd2_maxdo_6.14_i686-pc-linux-gnu
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 3 times, last edit by Sekerob at Dec 10, 2009 1:44:23 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
:D ok
data dir in oldie contains wcg_dddt_autodock_6.07_i686-pc-linux-gnu wcg_dddt_autodock_6.07_graphics_6.07_i686-pc-linux-gnu wcg_faah_autodock_6.07_i686-pc-linux-gnu +graphics wcg_flu1_autodock_6.04_i686-pc-linux-gnu +graphics wcg_hcc1_img_6.03_i686-pc-linux-gnu + graphics wcg_hcmd2_maxdo_6.14_i686-pc-linux-gnu + graphics wcg_hfcc_autodock_6.10_i686-pc-linux-gnu +graphics wcg_hpf2_rosetta_6.03_i686-pc-linux-gnu +graphics wcg_rice_6.17_i686-pc-linux-gnu +graphics wcgrid_cep1_6.32_i686-pc-linux-gnu +graphics data dir in quad has no apps.Reset new; message from server: resent lost result CMD2.... [error] App version has unsupported platform i686-pc-linux-gnu [error] No application found for task: i686-pc-linux-gnu 614 ; discarding any idea what the '614' is in the third quoted line? reset old; message from server: resent lost result HFCC_s1-0104.... [error] App version has unsupported platform i686-pc-linux-gnu [error] No application found for task: i686-pc-linux-gnu 610 ; discarding '610' on the old box ? no apps in old box anymore either.  this is fun, if Y'all had more problems like this i would participate here a lot more, i havent had this much fun 'cince docking alpha harmonious redundancy tests where my opteron was making up its own answers.  |
||
|
|
|