LHC@home

Hello!
Thank You all for the two latest flood of wu's, and at my computers everything has run without any hickups!
Just had a look at some of my pending wu's to see how things will end, when I just stumbled acrossthis work unit 1124180, one of the results have to be invalid take a look at result id 5800979 and my 5800981. It's a "lightyear" between the used time!!

Some body please explain this, both are valid according to the results! HMMMMMMMMMMMMMMm.

Thank You.

Edit.

If You take look at host Id 69468 it has a lot of "valid" results with 0 credit given!


And a lot of them ended up invalid, something seriously wrong with the pc?
With regards

Looking forward to the next batch of work!!

---

Hans Sveen
Oslo, Norway

Just loked at Unit 580979, that is strange one real fast and one a long time. But I have had a couple like that also lately, the validator waits till three match to see what is correct and grant credit.

One of mine like that work unit wjan1D_v6s4hvnom_mqx_nc__17__64.32_59.33__6_8__6__25_1_sixvf_boinc84507 had to wait for the 5th to be sent back to be validated, two real fast got no credit, the three longer running systems got credit. Did look strange with 4 returned and still pending.

Some units like that show why there are so many copies of each sent out, but should this be reissued? only the scientist can tell. Hard to believe that there would be that much diferance with the 2 systems that did not get credit.

Ray

---

Pizza@Home - Rays Place - Rays place Forums

Has anyone else had any of these? Looks almost like they should be reissued just to be sure that the results are correct. Would hate to have the real particals go wrong just because of a couple computers not giving good results. In some cases 3 out of 5 just may not be good enoughf.
But ones like these just may be what we are getting to download now for there reruns.
Ray

---

Pizza@Home - Rays Place - Rays place Forums

Another one unit 1124228 waiting for the 5th due to a wide varience between systems. That was from the little womans computer so it says Karens rather than my name.
Ray

---

Pizza@Home - Rays Place - Rays place Forums

Just loked at Unit 580979, that is strange one real fast and one a long time. But I have had a couple like that also lately, the validator waits till three match to see what is correct and grant credit.

One of mine like that work unit wjan1D_v6s4hvnom_mqx_nc__17__64.32_59.33__6_8__6__25_1_sixvf_boinc84507 had to wait for the 5th to be sent back to be validated, two real fast got no credit, the three longer running systems got credit. Did look strange with 4 returned and still pending.

Some units like that show why there are so many copies of each sent out, but should this be reissued? only the scientist can tell. Hard to believe that there would be that much diferance with the 2 systems that did not get credit.

Ray

Looks like one of the users has oc'd their host too far and is only computing for 2 seconds before bombing out. I'd hazard a guess that this might be the case for all the wu's where we are seeing fast completions and "normal" completions within the same wu.

Live long and crunch.

---

Paul
(S@H1 8888)
BOINC/SAH BETA

Looks like one of the users has oc'd their host too far and is only computing for 2 seconds before bombing out. I'd hazard a guess that this might be the case for all the wu's where we are seeing fast completions and "normal" completions within the same wu.

Live long and crunch.

That is what I was thinking at first, but was getting to many of them, than Hans found two in his pending units. With so many thought it may be something more synister, well a 50/50 chance of eather way I guess.

Ray

---

Pizza@Home - Rays Place - Rays place Forums

Looks like one of the users has oc'd their host too far and is only computing for 2 seconds before bombing out. I'd hazard a guess that this might be the case for all the wu's where we are seeing fast completions and "normal" completions within the same wu.

That is what I was thinking at first, but was getting to many of them, than Hans found two in his pending units. With so many thought it may be something more synister, well a 50/50 chance of eather way I guess.

Once the host has dumped the WU after two seconds, it's then free - so it
will download and trash another WU and so on. This means that a small number
of ropy nodes has a disproportionate effect on the overall success rate.
(This is sometimes known as the "black-hole effect" - all your jobs get sucked into a misconfigured node and your data never re-appears).

Even with 100 000 WUs in total, the 100 WU/day limit will let a bad host hit
0.1% of all of them - and if you look at how many WUs you deal with each day and how many active users there are, the chances of several people seeing them will be higher than you might expect.
{But I'm not going to embarrass myself by trying to work them out!}

Of course, that still doesn't tell you if it's a genuine problem with the machine or a deliberate ploy to gain credits by harvesting the half-totals LHC give out for invalid results.

Henry

Hi!

A couple of days ago one of my pc's (Id 107184) retuned some invalid wu's, I managed to suspend all communication thereby only created a small black hole!
The reason was my Amilo had some dust rabitts inside clogging my gpu fan. So I guess I'm not the only one with over heating.
So my advice is every pc needs frequent cleaning more frequent if You use a portable one!

Sorry to make it more difficult to have the credits flowing for some of You!

A couple of these wu's are already marked invalid, please also clean away the rest.

Still have a nice weekend and a happy crunching before this study runs dry; still about 100000 wu waiting!

---

Hans Sveen
Oslo, Norway