These is the list of
Top Computers

Hosts in positions 2) 239985 and 3) 128175 just appeared today (Anonymous users). At first sight it appears these might just be merged computers, but look further . . . these computers are downloading 1000s of w/u (just today 10-Nov many hundreds of w/u). Is it possible that this user is downloading daily quota of 50w/u --> dumping w/u --> reseating project --> merging hosts --> downloading w/u --> dumping w/u --> reseating etc etc in a long loop?

Is it possible for a computer to download 1000s of w/u every day using the above method? This could be a computer fault or maybe malicious reasons?

---

Did you notice how many of those WU were marked as "over"; "Client Error"; "Downloading"? I scrolled through between 5 and 10 pages of the first host and about 3 of the second host. Except for about the first 5 WU or so on each machine, the rest of them were download errors.


> These is the list of
> <a> href="http://setiweb.ssl.berkeley.edu/top_hosts.php?sort_by=total_credit">Top
> Computers[/url]
>
> Hosts in positions 2) <a> href="http://setiweb.ssl.berkeley.edu/show_host_detail.php?hostid=239985">239985[/url]
> and 3) <a> href="http://setiweb.ssl.berkeley.edu/show_host_detail.php?hostid=128175">128175[/url]
> just appeared today (Anonymous users). At first sight it appears these might
> just be merged computers, but look further . . . these computers are
> downloading 1000s of w/u (just today 10-Nov many hundreds of w/u). Is it
> possible that this user is downloading daily quota of 50w/u --> dumping w/u
> --> reseating project --> merging hosts --> downloading w/u -->
> dumping w/u --> reseating etc etc in a long loop?
>
> Is it possible for a computer to download 1000s of w/u every day using the
> above method? This could be a computer fault or maybe malicious reasons?
>
>

---

For the second computer EDIT (second of the two, this is the one in third place) EDIT almost all the 4400 units have this type of errors. Seems strange.

---

They are surely "outwoard" computers. But if using them the owner is also responible for the right function of it. We want to produce scientivic data here, not errors ;-). So have a look on your pc's.

---

WARR - Wissenschaftliche Arbeitsgemeinschaft für Raketentechnik und Raumfahrt
(WARR - scientific working group for rocket technology and space travel)

> For the second computer EDIT (second of the two, this is the one in third
> place) EDIT almost all the 4400 units have this type of errors. Seems
> strange.
>

What suprises me is the fact that he/she can download > 50 w/u x day. As far as I am aware download errors do count for the 50 w/u daily quota. Is this correct?

Now there are other interesting facts . . . computer in position 4) 115974 has only returned 32 results. How is it possible to have such a high total credit with just 32 results? When merging computers results also merge don't they?

---

> place) EDIT almost all the 4400 units have this type of errors. Seems

It is perplexing.
The daily download limit was implemented to prevent scenarios where a single host, that perhaps has an error, would deplete the server wu cache...

These hosts seem to have an error successfully obtaining wu's (Client Error - Downloading), but they are not affected by the daily download limit.

On the plus side, it looks like the owner(s) has determined that a problem existed and resolved it. I'd go as far as to suggest that it's the same owner as the errors stopped on both systems within 1 minute of each other.

If the owner(s) does not see this thread, perhaps one of the devs could contact them and query them on what went wrong. This only for information sake to try and determine why the systems managed to bypass the daily download limit, as there is IMHO something strange (strange wrong, not strange malicious) there.

---

> Now there are other interesting facts . . . computer in position 4) 115974 has
> only returned 32 results. How is it possible to have such a high total credit
> with just 32 results? When merging computers results also merge don't they?

Assimilated results are deleted, so this low number is not surprising. Given that it is a 2.4Ghz Quad CPU system created in late July, I'd say the credit is feasible. The RAC is very low, indicating that it is not presently running, thus accounting for the lack of any "in progress" wu's.

---

> > Now there are other interesting facts . . . computer in position 4)
> 115974 has
> > only returned 32 results. How is it possible to have such a high total
> credit
> > with just 32 results? When merging computers results also merge don't
> they?
>
> Assimilated results are deleted, so this low number is not surprising. Given
> that it is a 2.4Ghz Quad CPU system created in late July, I'd say the credit
> is feasible.
>

I didn't know this. Thank you Ralic

PS: I remember you! You are the wizard who saves us from the 4.12 machine locking. Thank you Ralic

---

Hardly a wizard. <*8^)
I just enjoy helping where I can :)

---

> Now there are other interesting facts . . . computer in position 4) 115974 has
> only returned 32 results. How is it possible to have such a high total credit
> with just 32 results? When merging computers results also merge don't they?
>
Keep in mind the results database does not contain all the results (see the database purged? thread on this forum).

Currently, one of my computers shows 21557 credits and only 123 results. That is an average of 175.3 credits per result on that system.

It will be difficult to make any sense of the whole credit system until the database is "put back together".

I'm not seeing how those computers can be granted so much credit if they are only returning errors...Also it blows my mind that they can complete so many WU's a day.

---

Kolch - Crunching for the BOINC@Australia team since July 2004.
Search for your own intelligence...

> I'm not seeing how those computers can be granted so much credit if they are
> only returning errors...Also it blows my mind that they can complete so many
> WU's a day.
>
On a day I did have problems with a box and I had to fire it up several times and had to reset and reattach to the project...
I never succeeded in getting more than the 50 WU's / day. I'd had to wait till the next day to get some WU's.
So this is definately VERY STRANGE.

GrtZ ;-))

---

Yes I hope BOINC developers do see this thread! These two machines are doing something that in theory should never happen. Can anyone provide or attempt an explanation?

---

> > Now there are other interesting facts . . . computer in position 4)
> 115974 has
> > only returned 32 results. How is it possible to have such a high total
> credit
> > with just 32 results? When merging computers results also merge don't
> they?
>
> Assimilated results are deleted, so this low number is not surprising. Given
> that it is a 2.4Ghz Quad CPU system created in late July, I'd say the credit
> is feasible. The RAC is very low, indicating that it is not presently running,
> thus accounting for the lack of any "in progress" wu's.
>

Sorry to correct the wizard, but the computer in in position 4 is not a Quad CPU system. It is a single 2.4GHz P4.

'Number of CPU's: 1'

---

<a><img src="http://boinc.mundayweb.com/stats.php?userID=525">

<a><img src="http://www.wombatradio.com/stats/rtb/sig.php">

> Sorry to correct the wizard, but the computer in in position 4 is not a Quad
> CPU system. It is a single 2.4GHz P4.
>
> 'Number of CPU's: 1'

LOL.
Everyone gets to be wrong sometimes....but this time it wasn't me :)
Have a look at the host number under discussion (115974). That host was in 4th place, but has now dropped to 5th.

I any event, you're also probably not wrong. For my money, I'd say it was a dual Xeon with hyperthreading turned on, giving the impression of a quad system...This was an afterthought of mine, but I didn't see the need to post it. Thought I'd do it now as I'm writing something anyway.

---