Today, 1/31/12 I got 2 more. http://einsteinathome.org/workunit/115182929, http://einsteinathome.org/workunit/115177415.
Since this problem was identified last week I would have thought the project would have deleted the bad data. IIRC, I remember I used to get messages deleting work for reasons unknown to me. It seems to be a waste of the projects resources allowing clients to process known bad data.
Yes, those two WUs have been canceled as a result of the discovery of more bad data. Unfortunately, it is difficult to to make sure (in advance) that only valid data is sent out to volunteers. So there may be occasions when tasks get pulled suddenly like this. Here is an excerpt of a message from Oliver that was sent to all Mods.
FYI, there's one additional rogue BRP dataset:
p2030.20111029.G36.97-00.93.N*
I'll cancel all related WUs/tasks over the next few hours (following a
kernel update on all E@H machines in Hannover)...
This message was sent around midday Hannover time on 31/1/2012. There was a followup message just after that to say that the action had been taken.
Quote:
Since this problem was identified last week I would have thought the project would have deleted the bad data.
Bad data is always deleted as soon as it is discovered. Unfortunately, this appears to be a new instance of the problem.
Quote:
IIRC, I remember I used to get messages deleting work for reasons unknown to me. It seems to be a waste of the projects resources allowing clients to process known bad data.
There is functionality within BOINC to allow the BOINC client to be advised when tasks on the client are no longer required by the server. A classic example would be a deadline miss task for which a canonical result already exists on the server. I'm sure there would be other situations as well. If such a task hasn't even been started by the client, it makes perfect sense to delete it. However, BOINC wont delete a task that is already underway.
For anyone particularly troubled by this ongoing problem, turning off FGRP tasks in your preferences would be a sensible course of action. There's no similar issue with GW tasks.
Go to you account page and follow the link to your E@H preferences. Click on the edit link and then deselect Gamma Ray Pulsar Search #1. Then click 'update' to save the change.
Today, 1/31/12 I got 2 more.
)
Today, 1/31/12 I got 2 more. http://einsteinathome.org/workunit/115182929, http://einsteinathome.org/workunit/115177415.
Since this problem was identified last week I would have thought the project would have deleted the bad data. IIRC, I remember I used to get messages deleting work for reasons unknown to me. It seems to be a waste of the projects resources allowing clients to process known bad data.
RE: Today, 1/31/12 I got 2
)
Yes, those two WUs have been canceled as a result of the discovery of more bad data. Unfortunately, it is difficult to to make sure (in advance) that only valid data is sent out to volunteers. So there may be occasions when tasks get pulled suddenly like this. Here is an excerpt of a message from Oliver that was sent to all Mods.
I'll cancel all related WUs/tasks over the next few hours (following a
kernel update on all E@H machines in Hannover)...
This message was sent around midday Hannover time on 31/1/2012. There was a followup message just after that to say that the action had been taken.
Bad data is always deleted as soon as it is discovered. Unfortunately, this appears to be a new instance of the problem.
There is functionality within BOINC to allow the BOINC client to be advised when tasks on the client are no longer required by the server. A classic example would be a deadline miss task for which a canonical result already exists on the server. I'm sure there would be other situations as well. If such a task hasn't even been started by the client, it makes perfect sense to delete it. However, BOINC wont delete a task that is already underway.
Cheers,
Gary.
And here is another
)
And here is another one.
http://einsteinathome.org/task/270492421
271717221 NG
)
271717221
NG
NG
Is this one of
)
Is this one of interest?
Task 272301472 in Workunit 116296456
RE: For anyone particularly
)
Well, how can I do that ?
RE: Well, how can I do that
)
Go to you account page and follow the link to your E@H preferences. Click on the edit link and then deselect Gamma Ray Pulsar Search #1. Then click 'update' to save the change.
Cheers,
Gary.
Thx - I got it now.
)
Thx - I got it now.
272579064 NG
)
272579064
NG
NG
273251777 NG
)
273251777
NG
NG