Hello,

for some reason I have two records for one machine in SETI database. When I try to merge them into one, the system does not present me with a table of computers to choose from. When I try to delete one of them, the system tells me that results exist for the computer I am trying to delete and refuses to do so.

Am I stuck? Do I require a manual intervention from sysadmins?

Regards,

-mm-

Hi Marek,

I have the same problem!
I cant delete it.

The one I can't delete has the same name but different computer...

Timmy

</img>

---

I have this problem also. I changed the OS on the box from ME to w2k box but I can't get them to merge.

If the computer has "enough" of a difference in the records they cannot be merged.

If they have any results, they cannot be deleted ...
<p>
For BOINC Documentation: Click Me!

There were no diference in the two machines hardware or OS wise. Other than I tried to process the WUs with the CLI version of BOINC client for a while, and this is when I noticed the extra entry.

-mm-

> There were no diference in the two machines hardware or OS wise. Other than I
> tried to process the WUs with the CLI version of BOINC client for a while, and
> this is when I noticed the extra entry.
>
> -mm-

Interesting, well it looks like the GUI and the CLI each thought that there was two different computers. WIth luck Rom will see this and look into this to see if that is a "difference" as far as the back end systems are concerned. You would not think so, but, it looks like that is the reason for the two records.
<p>
For BOINC Documentation: Click Me!

> There were no diference in the two machines hardware or OS wise. Other than I
> tried to process the WUs with the CLI version of BOINC client for a while, and
> this is when I noticed the extra entry.
>
> -mm-
>
>
There is a difference as far as BOINC is concerned. One is listed as having an athalon MP CPU and the other is not. That is enough difference to keep them from merging. Probably due to a change in the CPU recognition code recently.

I have seen threads suggesting ways of editing .xml files to get the computers into a state that they can be merged, but I don't recomend it. There seems to be too much chance of making a bigger problem if you try.

John Keck -- BOINCing since 2002/12/08 --

> There is a difference as far as BOINC is concerned. One is listed as having an
> athalon MP CPU and the other is not. That is enough difference to keep them
> from merging. Probably due to a change in the CPU recognition code recently.
>
> I have seen threads suggesting ways of editing .xml files to get the computers
> into a state that they can be merged, but I don't recomend it. There seems to
> be too much chance of making a bigger problem if you try.

Thanks John!

The master speaks! As always with the answer ... Heck I have problems just understanding the question ...

:)


<p>
For BOINC Documentation: Click Me!

Oh, but the Great Master - I have not touched the hardware. The same CPUs were in ther system before. The same CPUs are in the system now. Perhaps, what escaped your masterful attention, is that the back-end decided (erronously) to qualify the hardware as different. Ergo - with all the respect, o Great Master - you are, most likely, enjoying a logic-bug in the database design. BTW, same thing happened on LHC - something stininking in the code. I can smell it...

Yes, your Greatness, you have an answer - but is it a correct one? ;-)

-mm-

> Oh, but the Great Master - I have not touched the hardware. The same CPUs

That's not the point, the point is:

>> One is listed as having an athalon MP CPU and the other is not. That is
>> enough difference to keep them from merging. Probably due to a change in the
>> CPU recognition code recently.

It's not that the computer is different, it's that BOINC sees it differently.

> what escaped your masterful attention, is that the back-end decided
> (erronously) to qualify the hardware as different. Ergo - with all the

No, I think it's the core client that is reporting differently to the server. That, then, spawned a new ID for the system. On one hand, it might seem unfortunate, but it really doesn't make much of a difference in the long run, does it? WUs are still getting crunched and you're still getting credits for the work done.

That said, I do understand the frustration. In the last week, one of my systems managed to spawn not one but four new IDs. Yeah, I could merge them, but it's odd to see it happening for no apparent reason.

Cheers,

trane

> No, I think it's the core client that is reporting differently to the server.
> That, then, spawned a new ID for the system. On one hand, it might seem
> unfortunate, but it really doesn't make much of a difference in the long run,
> does it? WUs are still getting crunched and you're still getting credits for
> the work done.
>
> Cheers,
>
> trane
>

Allright! So, now that we seem to know what's causing the problem:

1. does anybody know how to fix it?
2. is anybody willing to fix it?
3. will it ever be fixed?
4. if it's going to be fixed, when approximately might that happen?

-mm-

p.s. no offence, but if the Sahara Desert was run using similar problem-resolution approach I have witness so many times in my short participation in BOINC - it would be out of sand in 48 hours, or so...

p.s. II please resist the urge to inform me that the folks in the back-room are working as fast as they can on solving some important WIN98 problem, or that crunching for the science is the overriding motif, or that the hardware is inadequate, or the like. I have heard this before, and don't care much for it...what I did not hear as of yet, is who really feels responsible for the broken pieces. Does anybody know?

These answers can be taken as scarcasm. They are not intended as such. You can choose to believe that or not. But they are an honest attempt to answer you question.

> Allright! So, now that we seem to know what's causing the problem:
>
> 1. does anybody know how to fix it?

Yes.

> 2. is anybody willing to fix it?

It is not broken.

> 3. will it ever be fixed?

If it is WAD, then there is nothing to fix ... (as a shameless plug, if you don't know what WAD means you have to look it up in the Glossary).

> 4. if it's going to be fixed, when approximately might that happen?

Never, see above.

> -mm-
>
> p.s. no offence, but if the Sahara Desert was run using similar
> problem-resolution approach I have witness so many times in my short
> participation in BOINC - it would be out of sand in 48 hours, or so...

Well, who needs that much sand anyway.

> p.s. II please resist the urge to inform me that the folks in the back-room
> are working as fast as they can on solving some important WIN98 problem, or
> that crunching for the science is the overriding motif, or that the hardware
> is inadequate, or the like. I have heard this before, and don't care much for
> it...what I did not hear as of yet, is who really feels responsible for the
> broken pieces. Does anybody know?

I feel responsible because as I was running SETI@Home Classic I wanted to be able to participate in many science projects because I like research. You might say I was a research "grouppie" because my mind does not work in a way that would allow me to do research, but I admire those who can create ...

I am always saddened by those, like you, that expect instant and perfect gratification. I am sorry that you don't like the answers that are given. Truth can set you free, but it is not always what we would wish it to be. And, yes, I am one of those that does say these things that you don't like. I feel bad about that too. And I too, wish it were otherwise. But it is not otherwise. We have what we have. We have problems with the system. They are being worked on, and over time, each and every day we have things just a little bit better.

As of right now, we have 3 projects that the bulk of us are participating in, SETI@Home and cp.net; and those of us that are lucky also have the Predictor@Home and LHC@Home to fill in where there are problems with SETI@Home. These projects are "streching" BOINC in different ways and showing places where there was not adequate thought, or just a plain old error. Perhaps as a developer myself I am more tolerant with a limited resource project as most of these are.

Being upset is not going to make anything better. Maybe this is not the time for you to be doing SETI@Home Powered by BOINC. In the mean time, might I suggest that you do either cp.net using either of the versions they have. Or you could go to Folding@Home that is not on BOINC (for now) and work those projects until BOINC is more mature ...
<p>
For BOINC Documentation: Click Me!

> > 1. does anybody know how to fix it?
>
> Yes.
>
> > 2. is anybody willing to fix it?
>
> It is not broken.

Dear Paul,

but it IS broken. Two records for the same exact hardware configuration is surely not what the developers intended.

Will you please be so kind and review the answers you provided to 1. and 2. above again? If somebody KNOWS how to fix it, the prerequisite of something being broken was satisfied. Therefore your answer to 2. can not be correct...

Regards,

-mm-

> > > 1. does anybody know how to fix it?
> >
> > Yes.
> >
> > > 2. is anybody willing to fix it?
> >
> > It is not broken.
>
> Dear Paul,
>
> but it IS broken. Two records for the same exact hardware configuration is
> surely not what the developers intended.
>
> Will you please be so kind and review the answers you provided to 1. and 2.
> above again? If somebody KNOWS how to fix it, the prerequisite of something
> being broken was satisfied. Therefore your answer to 2. can not be
> correct...
>
> Regards,
>
> -mm-
>
>
>
Marjerk, find another hobby !!!





Application has reported a 'Not My Fault' in module KRNL.EXE in line 0200:103F

Well said Paul and Steve!
L24

> Dear Paul,
>
> but it IS broken. Two records for the same exact hardware configuration is
> surely not what the developers intended.

No, it is NOT broken. The core client WAS broken (wasn't recognizing CPUs correctly) and then they FIXED it so it recognized your CPU for what it really is. At this point it saw a difference in hardware and created a new hostID. I have a couple computers listed in my account that are "ghosts". One from a major hardware upgrade and one from some instability during which I had to change FSB settings (no I wasn't overclocking :) I live and deal with it.


---------------------------------------
- A member of The Knights Who Say NI!
Possibly the best stats site in the universe:
http://boinc-kwsn.no-ip.info

Cut the blue wire.

> > > > 1. does anybody know how to fix it?> >
> >
> Marjerk, find another hobby !!!

Cold shot, Steve, but hysterically funny. :->

So, somebody got up on the back end, looked around and fixed the problem. I was able to merge the two accounts off a sudden. Both on LHC and SETI.

Despite all the rethorics of questionable quality, something was broken and something got fixed. It's a shame somebody did not communicate what this something was.

BTW, this is EXACTLY how user's trust in any computer system is destroyed. But apparently, somebody does not care.

-mm-

P.S. In case you are not certain, now would be an appropriate time to re-start your sorry insults...

---

* Wisdom is ability to get important things approximately right most of the time.
* Breakthrough is seeing what everyone else has seen, but thinking what no one else has thought.
* Cogito, ergo sum.

> So, somebody got up on the back end, looked around and fixed the problem. I
> was able to merge the two accounts off a sudden. Both on LHC and SETI.
>
> Despite all the rethorics of questionable quality, something was broken and
> something got fixed. It's a shame somebody did not communicate what this
> something was.
>
> BTW, this is EXACTLY how user's trust in any computer system is destroyed.
> But apparently, somebody does not care.
>
> -mm-
>
> P.S. In case you are not certain, now would be an appropriate time to
> re-start your sorry insults...
>


OK I will start , so you finally did it correctly huh!!!!;0)

---

Semper Eadem
So long Paul, it has been a hell of a ride.

Park your ego's, fire up the computers, Science YES, Credits No.