You have to wait for your quorum partners to return their copies of the work units.
Some of those machines may be in a week long back-off because of the recent scheduler problems. Other may just be machines carrying a large cache of work units.
Even if they're not return by the original computer that they were sent to, they'll be sent out again after they pass deadline.
You're credits aren't gone. They're simply "in the bank". Take a look at the BOINC FAQs or the Wiki to see how things work with credits.
And a week isn't all that long. I've got credits pending at VTU from the middle of December (and Sztaki is worse... people have pendings from last fall).
In the end, it's your decision what projects to crunch for. But no matter where you go, if there is a quorum system and you're the first to return a WU, then you'll have pending credits.
for results, that even been reported on 01/12!!!
I get no credits - so E@H gets no new cpu-time on my machine!
For this WU, http://einsteinathome.org/workunit/23645111, You gave yourself 9 days before returning result. From 12/30 to 01/08. So why won't You give others their "legal" 14 days to return results?
Agreed, given the initial replication is two, and all the backend trouble finishing up on S5R1, it shouldn't be too surprising there will be a bunch of pendings outstanding for a while yet.
It certainly is no reason to "punish" the project because of it. It's not like they can actually do anything about when the hosts run them or how long it will take to do them.
for results, that even been reported on 01/12!!!
I get no credits - so E@H gets no new cpu-time on my machine!
Actually, you probably should check your "facts" a bit more carefully. You have the grand total of 11 pending results out of 44 results total. The oldest pendings were sent to you on Jan 12 - you actually returned them on the 14th. That is only 5 days ago. There are still 7 days to go before the 14 day deadline expires.
There are results in your list which you have not returned for at least as long as this 5 day delay, if not longer. Should someone else be complaining about their pendings as well? - pendings caused by your actions?
This would just about have to take the prize for the most frivolous complaint I've seen in quite a while - particularly bearing in mind the problems all crunchers have had in communicating with the servers to receive and return work.
In fact, I'm more interested in hearing about whether Dave actually found a good reason why the backend went into meltdown mode in the first place.
Since it appears they eliminated other hardware problems after the inital trouble with the RAID, I'm wondering if it's something screwy with MySQL (which I assume they're using for the BOINC database).
pending pending pending pending pending pending ...
)
You have to wait for your quorum partners to return their copies of the work units.
Some of those machines may be in a week long back-off because of the recent scheduler problems. Other may just be machines carrying a large cache of work units.
Even if they're not return by the original computer that they were sent to, they'll be sent out again after they pass deadline.
You're credits aren't gone. They're simply "in the bank". Take a look at the BOINC FAQs or the Wiki to see how things work with credits.
And a week isn't all that long. I've got credits pending at VTU from the middle of December (and Sztaki is worse... people have pendings from last fall).
In the end, it's your decision what projects to crunch for. But no matter where you go, if there is a quorum system and you're the first to return a WU, then you'll have pending credits.
Kathryn :o)
Einstein@Home Moderator
RE: for results, that even
)
For this WU, http://einsteinathome.org/workunit/23645111, You gave yourself 9 days before returning result. From 12/30 to 01/08. So why won't You give others their "legal" 14 days to return results?
Agreed, given the initial
)
Agreed, given the initial replication is two, and all the backend trouble finishing up on S5R1, it shouldn't be too surprising there will be a bunch of pendings outstanding for a while yet.
It certainly is no reason to "punish" the project because of it. It's not like they can actually do anything about when the hosts run them or how long it will take to do them.
Alinator
RE: for results, that even
)
Actually, you probably should check your "facts" a bit more carefully. You have the grand total of 11 pending results out of 44 results total. The oldest pendings were sent to you on Jan 12 - you actually returned them on the 14th. That is only 5 days ago. There are still 7 days to go before the 14 day deadline expires.
There are results in your list which you have not returned for at least as long as this 5 day delay, if not longer. Should someone else be complaining about their pendings as well? - pendings caused by your actions?
This would just about have to take the prize for the most frivolous complaint I've seen in quite a while - particularly bearing in mind the problems all crunchers have had in communicating with the servers to receive and return work.
Cheers,
Gary.
Amen to that! :-) In fact,
)
Amen to that! :-)
In fact, I'm more interested in hearing about whether Dave actually found a good reason why the backend went into meltdown mode in the first place.
Since it appears they eliminated other hardware problems after the inital trouble with the RAID, I'm wondering if it's something screwy with MySQL (which I assume they're using for the BOINC database).
Alinator