Here is something I have never seen before:

http://setiathome.berkeley.edu/result.php?resultid=514987054

The report deadline for this result was 6 May 2007. I sent it up on 15 Apr 2007. The validate state is "Result was reported too late to validate"!??
Can anyone tell me why I didn't get credit for this result?
Thx,
Cobbler.

Normally it should have been granted credit, but there must have been some corruption in the database tables when the replica crashed last week.

There have been a few other reports of this happening last week, so you weren't alone.

HTH,

Alinator

At least it seems we have no problem with the process that delete the results. The result you posted is already gone :)

Regards,
Wander

---

At least it seems we have no problem with the process that delete the results. The result you posted is already gone :)

Regards,
Wander

Nice!! There goes another 67 credits. Luckily this project is all about science! :-)

Here is something that angers me a little bit more:
This is one of my rigs:
http://setiathome.berkeley.edu/results.php?hostid=2038018
As you can see (for a few more hours, maybe days) I seemingly detached my client. That's funny (NOT) because I have never ever detached any of my clients, and the rig is still happily crunching away on those very WUs.

I've got the feeling that all this is due to the possibility of the SETI team implementing the "initial replication 3 / minimum quorum 2" thing without properly testing it first.
Thx,
Cobbler

At least it seems we have no problem with the process that delete the results. The result you posted is already gone :)

Regards,
Wander

Nice!! There goes another 67 credits. Luckily this project is all about science! :-)

Here is something that angers me a little bit more:
This is one of my rigs:
http://setiathome.berkeley.edu/results.php?hostid=2038018
As you can see (for a few more hours, maybe days) I seemingly detached my client. That's funny (NOT) because I have never ever detached any of my clients, and the rig is still happily crunching away on those very WUs.

I've got the feeling that all this is due to the possibility of the SETI team implementing the "initial replication 3 / minimum quorum 2" thing without properly testing it first.
Thx,
Cobbler

Nice to see a positive attitude like yours :)

This "ghost detach" seem to be occurring more and more lately. I don't know if there is any relation, but the last major hiccup in the project was the failure in the replica Master Science DB.

If you have the time, check the WUs that are assigned to your host against those listed in the BOINC Mgr. A secondary effect of this problem is that even though the WUs are canceled with "Client Detached", the host somehow fails to recognize that and keep crunching the WU, and when it tries to report it in, the result is thrown away.

Regards,
Wander

---

...
I've got the feeling that all this is due to the possibility of the SETI team implementing the "initial replication 3 / minimum quorum 2" thing without properly testing it first.
Thx,
Cobbler

SETI Beta has been running that way for several months without any such problem. Granted the server load there is much less, but I can't think of a way they could have tested under full load.

The unique thing that happened here at about the same time was the zombie task cleanup running amok because the replica database was down. IMO that is the probable cause of these indicated unreal detaches.

                                                                  Joe

The unique thing that happened here at about the same time was the zombie task cleanup running amok because the replica database was down. IMO that is the probable cause of these indicated unreal detaches.

                                                                  Joe

I believe a large batch (possibly all my pending) of my SETI queue was marked as detached in the "results for computer" web page sometime in the last 24 hours. Already I've had one rejected for being two late (it was only two days old, and had many days yet to go to due date, so this case fits the "rejected if returned after a quorum from others" model).

If I'm right on timing, then this effect was continuing as recently as the wee hours Monday morning, Pacific time.

I'm running some Overclocking experiments on the machine in question, lots of reboots yesterday, and two (probably legitimate) errored results, so in my case it is not unthinkable that it was triggered by something on my computer.

In case another example is interesting, here is a link to the result page, for however long it stays up:

result 518665786

---

Yep, quorum had been set, so no joy. :-(

I don't think it had anything to do with your machine, mine isn't OC'ed and is showing the same thing.

Alinator