when Seti@home started up with the new XML file it downloaded everything except the new Seti@home application

setiathome_5.13_powerpc-apple-darwin

Hopefully the development team is working on this...

...and this:

SETI@home - Message from server: No work sent
SETI@home - Reason: no work from project

;)

---

It may not be 1984 but George Orwell sure did see the future . . .

Is the "computational error" being addressed by the SETI @home programmers ?

Hi to all,

I've the same problem (on a G5 PPC) since 3 days around I always get the message "Unrecoverable error for result......"

I've tried everything right now. RESTART, REINSTALL and and and... nothing works.

Has anyone an idea what could help?

---

This happened to me since this morning. The old workunits got
uploaded and Boinc is still crunching the remainig workunit.
But all the newly downloaded workunit are broken of with a
unrecoverable error and exit code 2. When I look at my results
page, there are now over 20 results with client errors.

When I look at some of the workunits, I can see that there are
more client errors from other Mac users. So it seems it's a
Mac related problem, but how can this be ? It can not be the
OS version, as 10.4.10 was installed within a week from release
and Boinc has been crunching since then without errors. So what's
the matter with Boinc or Seti@Home ??

My system:
PowerMac G4 2x 1.42 GHz
MacOS X 10.4.10
BOINC 5.4.9

---

Grt. Merlin

There was a problem at the 5.18 SETI application, but it looks like another problem with the new 5.21 application which should fix the G3 related problems described in earlier posts. I will write the dev a mail about this problem.

---

I've just upgraded the client to 5.10.7 on two of my computers and both of them are now giving me errors.
The MacBook Pro (= Intel cpu) says "Computation error" after crunching for a very short time and the G5 Mac Pro (= PPC cpu) says "Task (xxxxxxxxxx) exited with zero status but no 'finished' file. If this happens repeatedly you may need to reset the project."
I have now reset the project a few times on the G5 with no change.

Any thoughts?

OK - I've gotten the MacBook Pro to work by resetting the project a couple of times, but I can't get the G5 Mac Pro running. I've dropped down to v. 5.8.17 and tried a couple of the other versions available on the BOINC website, but all I get is the same message: "Task (xxxxxxxxxx) exited with zero status but no 'finished' file. If this happens repeatedly you may need to reset the project."

I'm at a loss now. Any suggestions?

I'm at a loss now. Any suggestions?

Ignore the message, if the WU would not be reset to 0 % computation.

---

I have looked a little bit closer to the different results, and it looks like there are different errors with differnt solutions :

Michael Sugermanns systems :
- 2982038 : Can't create shared memory: system shmat (5.18)
This means shared memory problems. Solution is to increase the shared memory
- 3078682 : Process creation (../../projects/setiathome.berkeley.edu/setiathome-5.21_AUTHORS) failed: errno=8 and "Maximum disk usage exceeded
Looks like to less diskspace assigned in the BOINC prefs. Please allocate more diskspace in the BOINC prefs and do a "reset project"

Merlin :
Process creation (../../projects/setiathome.berkeley.edu/setiathome-5.21_AUTHORS) failed: Error -1 execv: No such file or directory

Marcus72ch :
process exited with code 2 (0x2)

The Gonif :
process exited with code 8 (0x8, -248)


What did happen, if the systems which get the "exited with code ..." and "AUTHORS errors" would do a project reset ?

It looks there are only some systems are affected...

---

I have looked a little bit closer to the different results, and it looks like there are different errors with differnt solutions :

Merlin :
Process creation (../../projects/setiathome.berkeley.edu/setiathome-5.21_AUTHORS) failed: Error -1 execv: No such file or directory

What did happen, if the systems which get the "exited with code ..." and "AUTHORS errors" would do a project reset ?

It looks there are only some systems are affected...

I'll try this coming weekend after the current workunit has finished crunching.
Tonight I will lookup this AUTHORS file or directory, if this could be the cause
of my problem. Can you tell me if it is a file or a directory I should be looking
for ? Is this AUTHORS something new to Seti or Boinc ?

---

Grt. Merlin

I have looked a little bit closer to the different results, and it looks like there are different errors with differnt solutions :

Merlin :
Process creation (../../projects/setiathome.berkeley.edu/setiathome-5.21_AUTHORS) failed: Error -1 execv: No such file or directory

What did happen, if the systems which get the "exited with code ..." and "AUTHORS errors" would do a project reset ?

It looks there are only some systems are affected...

I'll try this coming weekend after the current workunit has finished crunching.
Tonight I will lookup this AUTHORS file or directory, if this could be the cause
of my problem. Can you tell me if it is a file or a directory I should be looking
for ? Is this AUTHORS something new to Seti or Boinc ?

It is a file. It contain GPL required informations. If the reset not work, I can put also on one my webserver. - I am not shure, if the file is not transmited, caused by a server or network problem.
Not shure, if it work to create an empty file in the projects/setistahome... directory. - Normaly the md5 checksums would not match and the BOINC client try to download it from the servers again.

---

Well, I've looked into the projects directory and the file
../../projects/setiathome.berkeley.edu/setiathome-5.21_AUTHORS
does exist and has an equal rights setting as the older file
../../projects/setiathome.berkeley.edu/setiathome-5.18_AUTHORS
The current workunit still uses 5.18 so the 5.21 seems to cause
my problems. The date and time of the AUTHORS file is set to
18-7 10:23 CET what coincides with the time the errors started
to occur. I'll move the 5.21 files out of the seti project
directory and try some new workunits ...


edit:
New workunit was requested and all the 5.21 files that I had
moved, were downloaded again to the same location ... but still
the Unrecoverable error with exit code 2 ... so how did you find
out that the AUTHORS file could not be found ? And if so, isn't
the 5.21 programm at fault since it cannot find the files that
were just downloaded ? I don't think a project reset will do
much good ...

---

Grt. Merlin

Well, I've looked into the projects directory and the file
../../projects/setiathome.berkeley.edu/setiathome-5.21_AUTHORS
does exist and has an equal rights setting as the older file
../../projects/setiathome.berkeley.edu/setiathome-5.18_AUTHORS
The current workunit still uses 5.18 so the 5.21 seems to cause
my problems. The date and time of the AUTHORS file is set to
18-7 10:23 CET what coincides with the time the errors started
to occur. I'll move the 5.21 files out of the seti project
directory and try some new workunits ...

Strange.
I think we have got to wait for the response from Eric (the developer) of the SETI application.

edit:
New workunit was requested and all the 5.21 files that I had
moved, were downloaded again to the same location ... but still
the Unrecoverable error with exit code 2 ... so how did you find
out that the AUTHORS file could not be found ? And if so, isn't
the 5.21 programm at fault since it cannot find the files that
were just downloaded ? I don't think a project reset will do
much good ...

I got this from the std err log from your WUs :

<core_client_version>5.4.9</core_client_version>
<message>
process exited with code 2 (0x2)
</message>
<stderr_txt>
2007-07-19 20:33:02 [SETI@home] Process creation (../../projects/setiathome.berkeley.edu/setiathome-5.21_AUTHORS) failed: Error -1
execv: No such file or directory

---

There is a new version (5.22) for MacOS PPC available.

---

Sorry, posted I think in wrong thread, will start new thread (I don't know that this is a "computational error").

There is a new version (5.22) for MacOS PPC available.

??? This version isn't available in the Boinc download section.
5.10.10 is the latest dev version there. Due to previous problems
with a 5.8 version, I'm reluctant to try a new version of Boinc.
But I'll try 5.10.7 (latest recommended version) after finishing
the current work unit ... I'll post here again after trying a newer
version of Boinc ...

---

Grt. Merlin

There is a new version (5.22) for MacOS PPC available.

??? This version isn't available in the Boinc download section.
5.10.10 is the latest dev version there. Due to previous problems
with a 5.8 version, I'm reluctant to try a new version of Boinc.
But I'll try 5.10.7 (latest recommended version) after finishing
the current work unit ... I'll post here again after trying a newer
version of Boinc ...

5.22 is the SETI application version. If you do a "reset project" there would be automaticly downloaded the newest version.

---

5.22 is the SETI application version. If you do a "reset project" there would be automaticly downloaded the newest version.

Ah, my mistake .. well, as I said, I'll wait for the current WU to end.

---

Grt. Merlin

There is a new version (5.22) for MacOS PPC available.

Ok, this versions seems to work. Just started on some new workunits
without errors. Hope it will stay that way. And I'll stick to the 5.4.9
Boinc version as well, no need to update it now...

---

Grt. Merlin