I'll post this here since the Beta forums aren't up yet.
Anyone else in Beta notice a high frequency of "Checked, but no consensus yet" WU's?

---

I'll post this here since the Beta forums aren't up yet.
Anyone else in Beta notice a high frequency of "Checked, but no consensus yet" WU's?

Yep.

---

https://youtu.be/iY57ErBkFFE

#Texit

Don't blame me, I voted for Johnson(L) in 2016.

Truth is dangerous... especially when it challenges those in power.

I'll post this here since the Beta forums aren't up yet.
Anyone else in Beta notice a high frequency of "Checked, but no consensus yet" WU's?

Yep.

Ditto :-)

---

I'll post this here since the Beta forums aren't up yet.
Anyone else in Beta notice a high frequency of "Checked, but no consensus yet" WU's?

Got 1 results so far. 5 others with no quoram ; 4 of which I am the only processor done so far. Processing alone is the feeling I get lol!

---

Got 1 results so far. 5 others with no quoram ; 4 of which I am the only processor done so far. Processing alone is the feeling I get lol!

Rest assured, you're not alone... 8)

---

Aloha, Uli

Can someone explain why I was given 0 credit for this wu
here. I think the wu completed successfully but being alpha maybe there was a problem I can not see. I don't mind no credit, just do not want to see this happen when it becomes a s@h app.

Can someone explain why I was given 0 credit for this wu
here. I think the wu completed successfully but being alpha maybe there was a problem I can not see. I don't mind no credit, just do not want to see this happen when it becomes a s@h app.

Your result did not match the other results - therefore it has a validate state of invalid. Now as to why it did not validate, I have no idea.

---

BOINC WIKI

Bones, it happened to me too: see here.
As long as that computer seems to be okay and is still able to produce results which are validated, I hope that even the results which get no credit make sense in a way.

---

@Finla, your result is also invalid, the same as mine. Click on the result ID link for that result to see the error. Now why it happened I'm yet to figure out. I'm concerned that something has gone wrong on my machine and as its also happened to you as well, it can not be an isolated incident. I've since had 1 result be granted normal credit but will keep an eye on others closely for problems.

Bones, Finla

I'm just curious. use explorer and open C:programfiles Boinc Slot Look in slot 0, 1, 2, 3 etc. Find the one with the seti 4.09, 4.11, 4.00, or 4.18 application that failed and look for messages like this one:

No heartbeat from core client for 30.994568 sec - exiting

do you both see this error?

[edit] actually I think only Bones will see this, Finla will probably see something else, or maybe they'll see nothing since moving on to a new WU

tony

Bones, Finla

I'm just curious. use explorer and open C:programfiles Boinc Slot Look in slot 0, 1, 2, 3 etc. Find the one with the seti 4.09, 4.11, 4.00, or 4.18 application that failed and look for messages like this one:

No heartbeat from core client for 30.994568 sec - exiting

do you both see this error?

[edit] actually I think only Bones will see this, Finla will probably see something else, or maybe they'll see nothing since moving on to a new WU

tony

Tony
What file would contain the message? I too am having similar errors to BOnes and Finla. While some are completing and getting credit, others appear to be ending early with client computing errors. I note that looking at the result on the web site they show a the heartbeat error you refer to. This problem appears on all of my machines. This is the error on the result sheet and appears on about 25-30% of WUs. All are windows systems XP2 both Intel and AMD processors, one of the computers is a dual CPU the others are single. Running multiple projects. Note message here since messages don't seem to be active over on BETA.

core_client_version>4.45
- exit code -1073741819 (0xc0000005)


No heartbeat from core client for 31 sec - exiting
No heartbeat from core client for 31 sec - exiting
No heartbeat from core client for 31 sec - exiting

Bob

---

No errors in the slot directory for me. I did find this in the stderrgui.txt file in the main boinc directory. The date the last error happened was the day my invalid wu completed so maybe that is what caused it. Hope this helps.

[edit] I have installed the symbol files now as I have read the sticky post relating to these type of errors and will see what happens

***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x0042F342 read attempt to address 0x00000010

1: 03/25/05 15:17:59



***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x0042F342 read attempt to address 0x00000010

1: 03/27/05 15:29:15



***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x00401560 read attempt to address 0x00000000

1: 03/28/05 08:31:01
1: c:boincsrcmainboinc_publicclientguiboincbaseview.cpp(137) +0 bytes (CBOINCBaseView::FireOnChar)



***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x0042F342 read attempt to address 0x00000010

1: 04/11/05 21:19:00


already tried both ports, giving up

***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x004190B9 read attempt to address 0x00000000

1: 06/11/05 15:58:14



***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x004190B9 read attempt to address 0x00000000

1: 06/19/05 08:57:57

I sent an email to Eric about the "Checked, but no consesus yet" that many of us run into. Here's what he answered me:

I've been noticing this is happening on Windows machines. The problem
is that some results are just not matching. Our quorum is 3 and
target is 4 continually. Our best guess at this point is that for
some reason, the BOINC client is occasionally unable to delete the
setiathome state files from previous runs and therefore starts a new
workunit in the middle.

Things shouldn't get stuck in this state, though. Results should be
sent out until a quorum is reached....

Will look into it.

Thanks,

Eric

---

Tony
What file would contain the message?
Bob

oops, should have mentioned it could be found in the Stderr file. I am getting the same error on Both Seti and Setibeta, but only one instance in each. for me Slot 0 is seti, slot 1 is seti/beta.

I'm just searching for a pattern and hoping some other guru will see a pattern if it exists and help fix it.

tony

[edit] I just rechecked and have 3 of those message on seti/beta

INFORMATIONAL MESSAGE FOR NON BETA USERS

I just looked and on my Athlonxp 2200+ Seti app 4.11(mavericks) is taking about 6 hours/WU and the 4.00 enhanced app (beta app) is taking about 12 hours.

I took a look into the regarding slot directory, and the stderr.txt was empty. So was stderrgui.txt in the BOINC folder.
In the meantime I got 2 more results which got 0 credit: wuid=48618 and wuid=49513

---

[/quote]
oops, should have mentioned it could be found in the Stderr file. I am getting the same error on Both Seti and Setibeta, but only one instance in each. for me Slot 0 is seti, slot 1 is seti/beta.

Thanks Tony, the Stderr files were empty. Also using variable slots on the different machines, for example one machine has seti/beta in 1 and seti in 2. Errors only occuring on seti/beta none showing on seti. Have not had new seti version come up, current version 4.09. In addition to heartbeat also seeing unhandled exception errors, the subject of another thread, also only on seti/beta. Just downloaded symbols file (per thread) to see if can catch something, here is example of error pre download,

core_client_version>4.45
- exit code -1073741674 (0xc0000096)



***UNHANDLED EXCEPTION****
Reason: Privileged Instruction (0xc0000096) at address 0x0023CAA4



Note running EStein, PP&H, CPDN, as well as Pirates (on Pirates have seen issue there with their starboard app not running properly, but that error hangs machine with failure to initialize, the others no problems. Like Finla the error above shows in the stderrgui.

Bob

---

Intel 3.4G P4 it takes aprox 4 hours.

GOOD NEWS TO LINUX USERS

but use at your own risk...as usual :)

Fortunately today Eric Korpela told me whereabouts of the source code of SETI enhanced (see this mailing list) and it's just in front of me. So I downloaded and built client for Linux. It requires some tweaks (some files are missing by cvs...) but it's now working apparently correctly.

So now here is the optimized statically linked binary for Linux (SSE2 enabled only, because all my libraries are P4 optimized) compiled with ICC9.0 and linked with fftw.

seti-enhanced-fftw-sse2-lin.tar.gz

And here is the source tree w/README how to build with ICC9.0. ICC 8.1 will work fine with minor change in the commpiler options. If you are interested in building it by yourself, it's easy. So anyone (especially Metod and Ned :) ) please build for other processors :) Just a moment...recently I installed Redhat9, so I can build for any other processors.....hmmm.....I'll build them.

seti-enhanced-fftw-src.tar.gz

My cruncher has crunched 5 WUs (including one reference) and it took 5000-5300 seconds/WU on my 2.8G P4 HT.....so average would be around 2500-2700 seconds. There is no "official" cruncher for Linux yet. But the development team is doing a good job...source code is much smarter than the previous (current) one.

A remaining problem is validation. I compared one result WU with the one the official Windows client had crunched, and found little difference. I hope it'll be within validation limits. But I haven't got any WUs validated yet (so far I'm the first one among the quorum.)

Next will be Window$ versions for all processors.....but just a moment, no, just a few days please...IMHO for the correct validation, I think I'd better make Windowss versions a bit later. If mine were wrong (it's unlikely....) and many people used mine (it's also unlikely), there would be a concensus between the wrong clients!!


and hey! add beta credits to the statistics!!


hmmmmmmm scared of blacklist....

---

Luckiest in the world. WMD = Weapon of Mass Distraction.
Click this table.

I will make clients for all processors both for Linux and Windows....so the names are

OSECL-X.tar.gz (Optimistic/Optimized Seti Enhanced Client for Linux)

and

OSECW-X.zip (ditto...for Windows)

where X is one of ICC's options in this page

The one mentioned above is OSECL-W.tar.gz

---

Luckiest in the world. WMD = Weapon of Mass Distraction.
Click this table.