1) Forums : Technical Support : Tasks will not complete (Message 11279)
Posted 20 Jun 2012 by Profile Conan
Post:
Running along the same lines are a few work units I have been getting on my 64 bit Fedora 16 Linux machine.
My Windows XP 32 bit machines do not seem to have this problem at all.

I have been getting a number of work units that have run way longer than most work units usually do.
Instead of say 14 to 16 hours which is the usual run times I have been doing they have run for as long as 32 hours or more. Plus they still get the standard 420 points for the effort.

I have just aborted two work units that appear to have restarted during their normal running time.
One had run for 40 hours, after 30 hours it was at 77% with 7 to go, this morning it was at 40 hours - 66% and 17 hours to go.
The other just stopped after 10 hours with 3 hours to go, at 77.496% for a number of minutes then restarted at 60% with 7 hours to go, so I aborted it as well.

This is only affecting my 64 bit Linux machine, very curious.
Just something else to think about.

I am posting this for information only as I have stopped work for the time being, so wont be doing any more work units.

Conan
2) Forums : General Topics : Work Units (Message 11260)
Posted 16 Jun 2012 by Profile Conan
Post:
Part of the problem, from what I have been reading from other users on other projects, is that the BOINC version you are using is the problem.

7.0.x does not balance out work loads like the older versions did.

7.0.x if it has a lot of work from one project wont download new work from another project even if they have the same work share amount (100 Rosetta, 100 Cosmology for example). You would have to wait until the project with a lot of work runs low or runs out, before it requests work from the next project.

If you suspend work for the other projects then you should get work for Cosmology.

This is how I understand it.

I am not using the latest BOINC at the moment as I could not get it to work on my Linux computer (some problem with identifying the version number or something), so did not worry with my Windows machines.

Conan
3) Forums : Announcements : Happy New Year! (Message 9791)
Posted 3 Jan 2012 by Profile Conan
Post:
And a Happy New Year to you to Faik and the rest of the C@Home Team.

May it bring great advancement in Cosmology thanks of course to the help of all us Volunteers.
4) Forums : Technical Support : Can't Seem to Modify Preferences (Message 9723)
Posted 3 Dec 2011 by Profile Conan
Post:
Can this preference issue please be looked at?
It has been going on for many, many months, I have not been able to update my signature due to the same error reported below.

I too have the wrong signature due to BoincStats making some database changes at least 6 months ago.

Conan
5) Forums : General Topics : Milestones (Message 9722)
Posted 3 Dec 2011 by Profile Conan
Post:
I didn't find a milestone thread so I started one as I reached 100k on Cosmology today.

Anthony.


You just had to look a little further back than you did.
See This Thread

I started the Milestone thread back in 2007 but the last entry was 1,047 days ago and I found it was me who last posted that I had reached 1,000,000 on the 20/1/2009.

Obviously no one had reached a milestone they wanted to crow about for some time, probably due to a number of problems with the project over that period.

Conan
6) Forums : Technical Support : Odd error occurred (Message 8974)
Posted 2 Jun 2010 by Profile Conan
Post:
These errors have been occurring on this project for over a year now and there has not been much of an attempt to fix them. If they have attempted to fix the problem then it has not worked.
Unfortunately we have not heard much from the project admins for quite some time.

The download error seems to lock up my Linux machines, so I had to stop. Windows doesn't lock up but for the time being I am doing other projects until I see some more life from this project.
7) Forums : Technical Support : Why can't i get a profile pic? (Message 8907)
Posted 4 Mar 2010 by Profile Conan
Post:
Everytime I try to upload a pic for my profile this warning comes up:

Fatal error: Call to undefined function imageCreateFromJPEG() in /home/boincadm/projects/cosmohome/html/user/create_profile.php on line 126

Anyone know what it means?


It means that a small part of the project software (Web programme or Boinc, I am not sure) is broken, as others are having the same problem.
Need Cosmology@home Project Admin to come along see this thread and then fix the problem.
Seems to affect new avatars or changes to old ones, if unchanged then avatar is still there.
8) Forums : Technical Support : URGENT PROBLEMS THREAD (2009 and after) (Message 8840)
Posted 22 Jan 2010 by Profile Conan
Post:
I'm new to all this . Can some one explain what is happening . I spend a lot of time crunching and am losing a lot of credits. This is the error message .
<core_client_version>6.10.18</core_client_version>
<![CDATA[
<message>
WU download error: couldn't get input files:
<file_xfer_error>
<file_name>params_011510_030623_1.ini</file_name>
<error_code>-200</error_code>
</file_xfer_error>

</message>
]]>
oldtroupador


You are not losing credits for this error.

It has been an ongoing issue at this project but has not been fixed in about a year now.
All that happens is that the download fails so that work unit never got downloaded so was not received at your computer and was not processed by your computer either.

If like me you run Linux it may cause an issue depending on your version of Boinc. I had Boinc stop working twice when the download failed and my Boinc Manager could no longer talk to the client and Boinc then stopped working (had to disable internet access abort the download and then reconnect the internet to get things going again).

On Windows I have not had this problem, still get the failed downloads (seems to be increasing) but no lock ups. I am using a later Boinc Client version on the Windows machine so that might be the reason for it not locking up.

So basically the failed downloads don't cause too much of an issue other than taking up bandwidth, it will then download a good one and you can keep crunching.
9) Forums : General Topics : CPU Time Less Than Elapsed Time (Message 8829)
Posted 18 Jan 2010 by Profile Conan
Post:
As this is Cosmology@home not Einstein@home, this should be posted on Einstein's message board for them to see it.

Have you noticed the problem on Cosmology?

I can't say that I have taken much notice of the CPU and elapsed times, I might just have to take a look.
10) Forums : Technical Support : ftp issues (Message 8822)
Posted 14 Jan 2010 by Profile Conan
Post:
At this stage there does not appear to be a 'solution'.

The problem has been going on for possibly up to a year now.
A number of other threads have been started about this issue.

The failed downloads recently caused my Boinc to stop working due to the corrupted download file. I had to disable network access then abort the download in order to get Boinc running again, it would start then Task manager would go blank saying it could not talk to Boinc client.

I connected a new computer to this project 2 days ago and out of 30 downloaded work units 8 had the problem you mention, that is nearly 27% failure on download.
It usually is not that high, seems the problem may be getting worse.
11) Forums : General Topics : Boinc User Of The Day (Message 8815)
Posted 10 Jan 2010 by Profile Conan
Post:
It has been an ongoing debate on various BOINC forums.
Users that have not contributed at all, they have just joined the project, can get User Of The Day.
It seems that the amount of work that you do is not the reason people get picked, it is just a random selection process.
I was told that if others vote for your profile then that can increase your chances of being selected.
Voting for yourself does not appear to work.
12) Forums : Technical Support : Too much time spent on a task with later report deadline (Message 8420)
Posted 8 Jun 2009 by Profile Conan
Post:
I have two active tasks, one with a report deadline of today, 6/7/09 9:26 AM and the other 6/16/09. I also run SETI@home, and with two CPUs, BOINC tends to run one Cosmology and one SETI task. But it has recently given preference to the later Cosmology task, spending 40 hours on it when it could have been working on the earlier one, whose deadline was passed this morning, having run for 39 hours with 89% progress.

I'm new to the Cosmology application, so maybe this is OK, but I'm tempted to discard the task.

What's the story here? Thanks.


With just two projects and an equal resource share, on a dual core machine the projects should run side by side, one project per core.

The other thing to consider is under preferences tasks will also switch at least every 60 minutes (unless you change this).

Now this would be fine if the Work Units were the same size, but Cosmology runs longer than standard Seti work units.

So what can happen is the Cosmo work unit stops after 60 minutes and probably another Seti work unit will start or the Seti work unit will stop and another Cosmo WU will start. Or the system can see the 50% preference split and not switch (but it usually does).

Either way if your system has been running for a while then you will see that 50% of your computers time is spread across each of your two projects.

However if one of your work units gets close to not meeting it's deadline (or Boinc calculates that it might not), then it will not switch from that work unit till it is complete.

This will then cause your computer to have an imbalance with the amount of time spent on each project (remember the other CPU core may still switch from one project to the other as per normal).
So then Boinc will only run the other project that did not have dead line problems for a while till the debts are balanced.

The balance is not immediate but will take some time to even itself out.
This may find you just doing Cosmo only for a while or Seti only for while.

Under normal circumstances no work unit should go over deadline unless the project has not calculated the correct runtime.
In which case Boinc has to then run a few work units to get a feel for how long they will run, and either stop downloading work or download more to meet your preference settings.
13) Forums : Technical Support : about claimed credit calculation.... (Message 8419)
Posted 8 Jun 2009 by Profile Conan
Post:
On my now older AMD Opteron 285 and 275 Linux machines ANY BOINC client that starts with a 6 gives me benchmark scores that can be at least 40% below what I am getting with 5.10.21.

I tried 6.4.7, 6.6.20 and possibly a third one but I can't recall the number.

They all killed my normal benchmark which has taken a large number of previous releases to get to that level.

Now they have dropped them all back again.

My same spec machine running Windows has had no change in it's benchmark when I installed 6.6.20 but had issues with one project so dropped back to 6.4.7 and all running OK.

My Windows machine benchmarks up to 20% higher than my Linux machines normally.
With the later Boinc Clients it goes to over 50% higher.
14) Forums : Technical Support : Download problems (errors?) (Message 8334)
Posted 10 May 2009 by Profile Conan
Post:
The error of 'not getting input files' is still happening.

It also seems to be on the increase, particularly on the 8th I had a lot of them, all failing to download correctly, I have had another couple since.
15) Forums : Technical Support : More Credit Needed (Message 8268)
Posted 29 Apr 2009 by Profile Conan
Post:
I think the project team got carried away. The news item said 5,000 credits for all ACTIVE users to compensate for problems.
However ALL users got the credit. Inactive users for months, years also got the points.

At least they are listening and are trying to help us so we can help them.
16) Forums : Technical Support : Longer / heavier WUs? (Message 8070)
Posted 10 Apr 2009 by Profile Conan
Post:
From what I can see is that it is the Windows application work units that are taking over 50% longer to do (I have one running for over 17 hours at the moment with 3 still to go), whilst the Linux work units on the same spec computer are taking 6 to 8 hours, like they use to.

Memory usage is about 176 MB in the early stage of the WU on Linux and going up to 350 MB in the later stages on both Windows and Linux.

The problem this brings up is awarded credit for these long running work units with the 140 limit that is now in place.
I have already had one claim for 225 credits and got 140 (9 cr/h) on the Windows machine.

I am wondering why have the Windows work units started to take up to 50% longer to run ???
Has only started with the new work units.


Well that 17 hour work unit ended after 24 hours, claimed 372.80 and got 140.00.


Memory usage has now shot up to between 540 MB to 587 MB at the 10 hour mark on two current work units.
Another now at 17 hours is using 247 MB.

I have 2 GB RAM but for the first time with Cosmology I am now running into a problem I sometimes get on The Lattice Project when I run more than 1 WU at a time ""Waiting For Memory"".

This is most unusual and my Linux machines still only take the normal times of 6 to 8 hours not 15 to 24 hours.
Will have to stop work on the Windows machine till this gets sorted as it is dropping to a crawl and I can't use it.


My last couple of Windows work units have taken from 18 to 24 hours, but now I have my first 2 Linux work units also going past normal times (one at 10 hours/41% done and one at 17 hours/86% done).

The two Linux work units start ""wu_040709_03014"", all other Linux types seem to run ok.
17) Forums : Technical Support : Longer / heavier WUs? (Message 8039)
Posted 9 Apr 2009 by Profile Conan
Post:
From what I can see is that it is the Windows application work units that are taking over 50% longer to do (I have one running for over 17 hours at the moment with 3 still to go), whilst the Linux work units on the same spec computer are taking 6 to 8 hours, like they use to.

Memory usage is about 176 MB in the early stage of the WU on Linux and going up to 350 MB in the later stages on both Windows and Linux.

The problem this brings up is awarded credit for these long running work units with the 140 limit that is now in place.
I have already had one claim for 225 credits and got 140 (9 cr/h) on the Windows machine.

I am wondering why have the Windows work units started to take up to 50% longer to run ???
Has only started with the new work units.


Well that 17 hour work unit ended after 24 hours, claimed 372.80 and got 140.00.

Memory usage has now shot up to between 540 MB to 587 MB at the 10 hour mark on two current work units.
Another now at 17 hours is using 247 MB.

I have 2 GB RAM but for the first time with Cosmology I am now running into a problem I sometimes get on The Lattice Project when I run more than 1 WU at a time ""Waiting For Memory"".

This is most unusual and my Linux machines still only take the normal times of 6 to 8 hours not 15 to 24 hours.
Will have to stop work on the Windows machine till this gets sorted as it is dropping to a crawl and i can't use it.
18) Forums : Technical Support : Longer / heavier WUs? (Message 8031)
Posted 8 Apr 2009 by Profile Conan
Post:
From what I can see is that it is the Windows application work units that are taking over 50% longer to do (I have one running for over 17 hours at the moment with 3 still to go), whilst the Linux work units on the same spec computer are taking 6 to 8 hours, like they use to.

Memory usage is about 176 MB in the early stage of the WU on Linux and going up to 350 MB in the later stages on both Windows and Linux.

The problem this brings up is awarded credit for these long running work units with the 140 limit that is now in place.
I have already had one claim for 225 credits and got 140 (9 cr/h) on the Windows machine.

I am wondering why have the Windows work units started to take up to 50% longer to run ???
Has only started with the new work units.
19) Forums : General Topics : Cosmology is back ? (Message 7968)
Posted 3 Apr 2009 by Profile Conan
Post:
Thanks Cosmo Team for getting this project back up an running.
Only lost 3 work units to a validation problem, was able to get credit for the rest.
I lost all my work when Einstein went down for just 3 days.

So well done there.

Nice addition to the Host data with the links to the stat sites.

Now let the work commence.
20) Forums : Technical Support : URGENT Problems Thread (Message 7929)
Posted 9 Mar 2009 by Profile Conan
Post:
The newest WUs I got were very strange:
12787402, 12787400, 12787399, 12787372, 12787357

They finished after about 500 sec, got validated and received 360 credits each.
Something went incredibly wrong here. Cosmo was always granting too many credits, but this is outrageous.


Get over it Saenger,
They are TEST work units for a reason, to TEST. So they may go for 500 seconds or they may go for hours, or they may not go at all and then crash. You are being paid for the fact that these work units may go pear shaped and the project is thanking you for running them.

Get off the credit thing. The project informed us what they were doing. If you don\'t want these \"outrageous\" work units then turn off \'run test work units\' in your preferences.


Next 20