boinc reports, the bit my cpu has to compute will take 294023 h. Will my mashine last this long? Will my pump last this long?
Windows xp sp2
How shall i behave?

Goodness! Never saw one that far removed from the mainstream before. I can only guess that the same Time Steps are being processed over and over again, but the run-time keeps accumulating, and boinc makes that wild (wilder than usual) calculation. In your Preferences, is the \"leave in memory while running\" option set to \"yes\"? (It should be, for best throughput for the Model.)

System date shouldn\'t make a difference, even if the CMOS battery died and the date reverted to system default. For my curiosity, is the system date correct?

As to what you should do, ignore it -- particularly if CPDN is the only boinc Project you run. The only problem is that boinc will think the Run can\'t be completed in time and will keep it going, at the expense of other boinc Projects, trying to fit it in. (It won\'t take that long. My guess is about 2500 hours for your machine.)

Edit: This is from the error report for one of your crashed Models (another Model crashed with \"Negative Pressure\" calculation):

CPDN Monitor - Quit request from BOINC...
set_worker_timer(): timer thread not working

That\'s an error I\'ve never seen before but it makes me suspicious that there is a problem with your machine or boinc has become corrupted. Is the machine over-clocked?
It would be easy enough to eliminate a boinc installation problem by re-installing boinc in the existing folder. That won\'t hurt your Model\'s folders but it assumes that the boinc control files aren\'t corrupted.

If the machine is overclocked, running Prime95 Torture Test for 24 hours would test stability.
Prime95 available here: http://www.mersenne.org/freesoft.htm

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.

Do you think this might be 2940.23h rather than 294023h ? The s/ts time isn\'t too far out from normal, as astroWX says 2,500 hours would be about what is expected. Perhaps some of the computer\'s statistics are messed up.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

believe me, Mike, the actual \"to completion\" is 291771:11:02 (progress is fast, isnt it?)
I don\'t see, where to chose \"leave in memory while running\" in my preferences (no such choice in /advanced/options)
The mainboard and cpu (and so cmos battery) is just 6 months old, i have not experienced any errors.
system date and time are correct and have always been.
The system is not overclocked and runs stable.
I think I\'ll practice ingorance and watch the fast progress?!
Thanks, georg.

The 3 preferences are on the server in your Account page. See \"Your account\" in the blue menu to the left for one way to get to them.

Do you mean \"Leave applications in memory while suspended?
(suspended applications will consume swap space if \'yes\')\"?

Then the answer is \"yes\"

Communicated with Thyme Lawn and he doesn\'t have a specific recommendation other than to contact Carl or other staff, which I did. (Edit: My primary interest in doing so is the timer thread issue.)

Georg,
Yes, that\'s the parameter I meant. (The \"Yes\" closes another door...)

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.

My guess from experience on other projects is that the RDCF (\'Result Duration Correction Factor\') is messed up.

Georg, could you go to your computer summary page on this website (\'Your account\', view computers on this account, click on computer ID for the computer in question), and tell us what the RDCF is? (almost the last value on the page, just above the choice for location: you can see the figure, but we can\'t).

Normally, you\'d expect to see something close to 1: yours is possibly much higher, even 100 or more.

The general advice is \"leave it alone, it\'ll sort itself out after about 30 WUs\" - which is going to take some time on CPDN! It can be fixed by getting down\'n\'dirty with a text editor on client_state.xml, but that would only be recommended for experienced programmers and system tweakers. Not worth doing at this stage unless you really want to clean up your display - better just to sit and watch, as you\'re planning to do. [It may cause problems when you\'ve finished this model and want to download another - that would be the time to sort it out, if at all].

I think RDCF errors happen if a benchmark goes wrong - I did it once by re-booting a computer during a benchmark run - but there may be other causes.

Result duration correction factor 176.639894
Now we seem to see clear.
I will not freak out, if my display is not clean.
and it\'s just a 290537 h left!

It\'ll be harmless until the current model is finished, but then you could have real trouble getting another. If that happens, \'reset\' project may fix it, or failing that, detach/attach (but don\'t do either reset or detach while you have a running model).

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

new question that might be connected to the problem that has sufficiently been diskussed (that\'s why i reopen this thread):
On results of my computer I get the report, that I have finished 6 WUs, which all ended with \"Client error\".
Is this normal?

Not normal. The machine has a variety of errors hinting at instability. Again, 24 hours running Prime95 Torture Test would tell the tale on stability.
http://www.mersenne.org/freesoft.htm

[Edited for typo.]

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.