Just installed 4.43, all went smoothly, but had to reboot to get the version change detected. Another program at my login is conflicting with Boinc because I do have problems getting it to run on its own. It takes closing Boinc and restarting it to do the trick, so its only really a problem if auto Windows update restarts the computer and I'm not there to restart Boinc. Of course I consider that to be losing valuable crunch time! ;)

---

Are people still seeing the v4.43 core client spike to 99% CPU utilization even when benchmarks aren't running? If this bug is really confirmed, it would be a pretty big reason for sticking with v4.25 or v4.19.

It has not happened (again) in the past 48 hours - I hope it was just co-incidence that I looked at task mamanger as BOINC was running the Benchmark.

BOINC.EXE on all three systems hasn't clocked up more than 2 minutes total CPU time in 48 hours whic is fine.

I'll keep checking it when I remember. I wish I could think of a way of setting an alarm to myself if the science app drops below say 95% for 5 minutes or so.
I'll have to see if I can write somthing to do this.

---

How long does 4.43's scheduler take to settle down? On the one machine I've upgraded, once it decided that it was able to download anything, instead of downloading more seti work, it went off and downloaded another CPDN workunit, now, ordinarily this wouldn't be problem on a modern fast computer, but, mine isn't, the current estimate for completion of it's current climate unit is slightly over 1600hours (still within deadline), there is no way, however, that the new unit will finish on time, so, what has boinc done? I'm sure that if I dump this workunit, then it will download another, or penalize me somehow, is this assumption correct? Any workarounds?

Thanks in advance for any help

---

Ken Phillips

BOINC question? Look here



"The beginning is the most important part of the work." - Plato

Any workarounds?

Sure. Do what those using any dev versions with the new scheduler have mostly decided to do and select the no new work function. That works fine.

---

Kolch - Crunching for the BOINC@Australia team since July 2004.
Search for your own intelligence...

Some weeks ago, I make a try with 4.25 version, after some days a got a error message "Connection failed", I came back to v. 4.19, everything was fine at this point. Some days ago I decide to give another chance to the newest versions and I install v. 4.38, for the first days everything seems fine, but the "dark side" strikes again and the message came back. I switch again to the 4.19 version that I still have in action. Understand that I'm little bit frightened of downloading and install v4.43 1 Is this version free of this issue ? Thanks for answering me !

---

How long does 4.43's scheduler take to settle down? On the one machine I've upgraded, once it decided that it was able to download anything, instead of downloading more seti work, it went off and downloaded another CPDN workunit, now, ordinarily this wouldn't be problem on a modern fast computer, but, mine isn't, the current estimate for completion of it's current climate unit is slightly over 1600hours (still within deadline), there is no way, however, that the new unit will finish on time, so, what has boinc done? I'm sure that if I dump this workunit, then it will download another, or penalize me somehow, is this assumption correct? Any workarounds?

Thanks in advance for any help

I thought that I had this one fixed.

How many CPUs including hyperthreading do you have?
What are the resource shares for your projects?
Do you know what the LT debt of CPDN was at the time that a new WU was downloaded? If so, what was it?

---

BOINC WIKI

I thought that I had this one fixed.

How many CPUs including hyperthreading do you have?
What are the resource shares for your projects?
Do you know what the LT debt of CPDN was at the time that a new WU was downloaded? If so, what was it?

Thanks for your replies, I have one cpu (AMD athlon 900 ish), no hyperthreading, resource share 33.33% each for 3 projects (LHC, CPDN, Seti), I have seen this about LT debt being bandied around a bit since these newer 4.3x clients have been released, but haven't taken the plunge till now, I haven't the faintest idea where to find the number at the moment, so unfortunately I can't tell you that. :-( If I find out before my edit time expires I'll post it to this message.

I'll implement the no new work thing for now, as otherwise I suspect that this other CPDN unit lurking in the background might adversely affect boincs' 'thought' processes, re. the other projects.

TTFN

---

Ken Phillips

BOINC question? Look here



"The beginning is the most important part of the work." - Plato

I thought that I had this one fixed.

How many CPUs including hyperthreading do you have?
What are the resource shares for your projects?
Do you know what the LT debt of CPDN was at the time that a new WU was downloaded? If so, what was it?

Thanks for your replies, I have one cpu (AMD athlon 900 ish), no hyperthreading, resource share 33.33% each for 3 projects (LHC, CPDN, Seti), I have seen this about LT debt being bandied around a bit since these newer 4.3x clients have been released, but haven't taken the plunge till now, I haven't the faintest idea where to find the number at the moment, so unfortunately I can't tell you that. :-( If I find out before my edit time expires I'll post it to this message.

I'll implement the no new work thing for now, as otherwise I suspect that this other CPDN unit lurking in the background might adversely affect boincs' 'thought' processes, re. the other projects.

TTFN

client_state.xml look for ling_term_debt

---

BOINC WIKI

I installed 4.43 after uninstalling the earlier 4.35 version. First time Boinc Manager started I had the "not connected to local host" situation. After closing Boinc Manager and restarting it no problems, but something strange: using HT I normally have 2 applications running. When looking in taskmanager I had 3 S@H applications but one of them had no CPU time. I closed Boinc Manager but still had the one S@H application with no CPU time. When trying to close it within taskmanager I couldn't succeed to close that one.

Rebooting the host, starting manually Boinc Manager: no problem. Closing it and restarting again, no problem.

Maybe the fact I didn't reboot after the installation was the cause of the 2 mentionned behaviors.

Maybe the fact I didn't reboot after the installation was the cause of the 2 mentionned behaviors.

Hi Thierry, I too had different "connection failed" issues, the client behaved differently when I restarted and then did a hard boot. I kept getting the "Connection failed" message through two or 3 boots on two machines. Now, for some reason, neither gives me that message now. Strange.

client_state.xml look for ling_term_debt

Cheers John,

If I understand the structure of that file correctly, the LT debt figures are:-
Climate prediction "-1788.269297"
LHC "4796.520069"
Seti "-3008.250772"

Lhc (as you probably well know) is out of work for the present.

Oops! Nearly forgot to mention, preferences are set to leave in memory, and connect every two days.

---

Ken Phillips

BOINC question? Look here



"The beginning is the most important part of the work." - Plato

client_state.xml look for ling_term_debt

Cheers John,

If I understand the structure of that file correctly, the LT debt figures are:-
Climate prediction "-1788.269297"
LHC "4796.520069"
Seti "-3008.250772"

Lhc (as you probably well know) is out of work for the present.

The CPDN debt is less than half an hour negative, and very well could have been positive when the request was made. How much time was left on the old CPDN WU?

---

BOINC WIKI

The CPDN debt is less than half an hour negative, and very well could have been positive when the request was made. How much time was left on the old CPDN WU?

It had (still has) just over 1600 hours to go (almost an eternity) good job the deadlines are long :-)

I actually did the update to 4.43 3 hours ago, the debt figures are from about 10 minutes ago 19:22UTC.

Forgot to mention, preferences are set to leave in memory, and connect every two days.

---

Ken Phillips

BOINC question? Look here



"The beginning is the most important part of the work." - Plato

Hi Thierry, I too had different "connection failed" issues, the client behaved differently when I restarted and then did a hard boot. I kept getting the "Connection failed" message through two or 3 boots on two machines. Now, for some reason, neither gives me that message now. Strange.

Hello Tony,

Since v4.25 I didn't use anymore the autostart of Boinc when rebooting even when I changed to 4.30 and then to 4.35. Same for now with 4.43.

What is nice is the fact Boinc Manager shuts down very quickly when closing it. With 4.25 to 4.35 it took some 10 to 15 seconds.
Now with 4.35 it's nearly immediately.

Still a pity the sort function we had in the earlier versions before 4.25 (or was it 4.19?) is still not build in.

The CPDN debt is less than half an hour negative, and very well could have been positive when the request was made. How much time was left on the old CPDN WU?

It had (still has) just over 1600 hours to go (almost an eternity) good job the deadlines are long :-)

I actually did the update to 4.43 3 hours ago, the debt figures are from about 10 minutes ago 19:22UTC.

Forgot to mention, preferences are set to leave in memory, and connect every two days.

I need to check what was actually built into 4.43, this should not be happening. The only times that a second CPDN WU should be downloaded are:
1) An idle CPU.
2) If the CPDN WU that you have is nearly completed.

---

BOINC WIKI

I need to check what was actually built into 4.43, this should not be happening. The only times that a second CPDN WU should be downloaded are:
1) An idle CPU.
2) If the CPDN WU that you have is nearly completed.

Cheers John, FYI (if it works) the link below should get you my complete message file from the time of the update to my present time (21:17 UTC)
update from
4.56 to 4.43 messages
Yes, I really was running 4.56 on this host, I liked it, it never let me down so I stuck with it, then got very confused when the version numbers went down again, so I left it to get on with the job, instead of tampering :-)

TTFN

---

Ken Phillips

BOINC question? Look here



"The beginning is the most important part of the work." - Plato

I need to check what was actually built into 4.43, this should not be happening. The only times that a second CPDN WU should be downloaded are:
1) An idle CPU.
2) If the CPDN WU that you have is nearly completed.

Cheers John, FYI (if it works) the link below should get you my complete message file from the time of the update to my present time (21:17 UTC)
update from
4.56 to 4.43 messages
Yes, I really was running 4.56 on this host, I liked it, it never let me down so I stuck with it, then got very confused when the version numbers went down again, so I left it to get on with the job, instead of tampering :-)

TTFN

I need a username and password. Email me at jm7 at acm dot org.

---

BOINC WIKI

did you download 4.43 after it became the recommended version or are you using 4.43 development version? (don't know if there is a difference)
i had the problem with the client downloading another WU every time i restarted the computer. only running Einstein, "connect to" = 0.04 , over 10H remaining on current WU. had this problem on 4.42, 4.43(dev). before i un installed 4.43(dev) i did file/suspend. installed 4.43(rec) and un suspended BOINC and BOINC downloaded a new WU. haven't been able to provoke this behavior after that. i aborted the extra WU so i only had the running WU and rebooted, but it didn't provoke it to download a new WU.

---

I've installed 4.43 on three computers with no problems noted so far except the Statistics Tab. Anyone else having a problem getting the stats to show up?

---

http://www.teamstarfire.org/

Had to go back to using 4.35 because 4.43 :

1) takes over cpu, stalling progress of the wu
2) causes errors for burp wu
3) generates ghost wu

On my system 4.35 runs very well (my 4.35 is the pre release version, the 4.32/4.35 combo) This was the only development version that runs properly on my computer.

---

98SE XP2500+ @ 2.1 GHz Boinc v5.8.8

And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer.