Client Error
3 Dec 2008 3:45:41 UTC
Topic 194068
(moderation:
Change that to server error.
No way do all my pc's fail....and only with Einsteins.
They run LHC and even that waste of time seti with no problem.
Now even my dual is doing 2 good ones and then a page full of nothingness over the event horizon.
BAH
computation error.....computation error.....computation error.....
Client Error
)
Interesting... You're running the _1 app, which is SSE, on systems that are SSE2-capable. The only system you have that can't do SSE2 is your Athlon XP. Try using either the 6.05 or 6.07 apps on one of the hosts and see if the issue goes away...
Hi, Please also check if
)
Hi,
Please also check if you are using the CPU throttling preferences setting in the Web GUI (use at most xxx % of CPU). There's a known bug in BOINC that can lead to early app aborts with this setting set to anything other than 100%.
CU
Bikeman
I only ask because it seems
)
I only ask because it seems odd to me that this computer was created a few days before the last time it contacted Einstein, but it also has had no further contact since 20 Nov 2008 5:21:09 UTC and then this computer started getting work again at 20 Nov 2008 8:35:49 UTC. Are they two seperate computers? I see one has Boinc version 6.2.19 and the other 5.8.16 so I'm guessing they are.
If they are the same computer there's a good chance something happend and I would do some test on the pc. Also upgrading the pc's with Boinc versions 5.8.16 and 5.10.30 might be worth looking into to.
On this computer I think it's what Bikeman mentioned.
RE: Also upgrading the pc's
)
I run 5.8.16 and have not had problems... Unless it's the CPU setting, I don't think it's a BOINC problem...
RE: I run 5.8.16 and have
)
I wasn't suggesting it because I thought it was a Boinc version issue. I suggested it becuase I thought his current versions might have corrupted files, and possibly explain the errors.
I think I used 5.8.16 before going to 5.8.18 for a while myself, and only swithed to 5.10.45 because it had a couple of extra options, one of them was the Advanced menu I believe.
Thanks for taking a look at
)
Thanks for taking a look at this guys.
I didn't notice anyone checking so I didn't look that last couple days here.
There are 3 pc's on the list that are actually here at home.
The one named "HAL 5767" is a dual P4 3.2 running 6.04 amd it runs with no problems for about 2 years.
One named "HAL 3000" is a AMD 3200 running 6.04 and it also runs with no problems for about 3 years.
Preferences all set at 100% and they run 24/7
Now it started saying there is 2 named "MAGIC2" (P4 2.50) with that slight difference now saying [fpu tsc sse sse2 mmx] but these are BOTH the same pc which I have been running here since I started here 4 years ago.
And it ran LHC the last 4yrs and seti classic before that.
So this is the one I have the problem with running Einstein.
You can see it ran that older version since that version but the other day it started having problems running them and some were finished 100% but then failed when they were being sent and also loading started failing and communication problems and I watched it happen so I uninstalled and reinstalled the newer version and the same problem continued so I uninstalled and reinstalled the older version and it would not even start running most of the time or when it finished it did the "client error"
The thing is it still runs LHC and the current version of seti no problem 24/7
Mainly seti since you know how LHC runs.
Preferences always 100%
I finally just gave up and uninstalled Einstein and have just been letting it run seti and LHC and it hasn't had any problem running them.
Just the Einstein no matter which version (except the apps/beta's)
Do you have any idea why it changed to saying [fpu tsc sse sse2 mmx]??
The only changes I have done to any of these is a security system update and it made no difference with the other 2 machines here (and I shut down "butch racer" and started running the newer dual D CPU 2.80GHz in it's place and it runs fine also)
Thanks
-Samson-
Seems there's been alot going
)
Seems there's been alot going on with that pc, so it's hard to say just what the problem is for Einstein and why it's not affecting the other projects.
I see from the errors you're gettng exit code -1 (0xffffffff) which was explained here:
Could be a place to start trying things out.
Since you've had problems undating Boinc it could be corrutped files in the program are in play.
Since there was also a security update, this may have a bearing on it as well.
I'd run chkdsk /f and see if it's not a system file problem, as well as testing the memory if more simple things don't correct the problem.
RE: Now it started saying
)
This honestly sounds a lot like the broken update to 5.10.28 that I tried from 5.8.16 and is one of the reasons I refuse to mess around with the BOINC Client until I'm forced to... I know I said that I didn't think it was a BOINC problem, but now that you've described things more it does...
I would run down all work you have on that system, then completely uninstall BOINC, then install it again and attach again. You may get duplicate host IDs, but if the computer is named the same, then you can now merge them by the host name. That is what I had to do when the installation for 5.10.28 blew up on me...
Even worse now. When I
)
Even worse now.
When I first brought this up it was just on my P4 2.5 machine and I did several "resets" with no luck and a few complete uninstall/reinstall erasing anything that said "boinc" and I got it to work for a month but now back to either just running Seti-Astro's or Einstein Arecibo Binary Radio Pulsar with NO PROBLEM but S5R5's FAIL over and over "client error" and of course after that I have to wait 24hrs to get more work hoping to get a Arecibo Binary Radio Pulsar instead since they work every time.
BUT NOW my dual processor P4 3.2 is doing the same thing even after "resets" and complete reinstalls of the current boinc version "6.6.20 for Windows"
Makes no sense and it sure did make my RAC drop from 1800 to 1300
And since it is just pure luck hoping to get any " Arecibo Binary" to do these 2 machines just sit here wasting time on the Seti-Astro's for a couple hundred hours.
How can this be a "client error"??
I have my other AMD 3200 sitting between these 2 machines and it has no problem doing any type.
RE: How can this be a
)
How can it not, when all work units I checked got completed successfully by your wingmen?
And it is quite annoying to read posts that are shouted with bold, red and size 16. So, I refrain from further suggestions to resolving your problems.
Gruß,
Gundolf
Computer sind nicht alles im Leben. (Kleiner Scherz)
