have just uninstalled 6.12.34 and installed 6.12.33 as advised. Here is a copy/paste of the beginning of the log files from the 6.12.33 client:


22/09/2011 22:59:43 | | Starting BOINC client version 6.12.33 for windows_x86_64
22/09/2011 22:59:43 | | log flags: file_xfer, sched_ops, task
22/09/2011 22:59:43 | | Libraries: libcurl/7.19.7 OpenSSL/0.9.8l zlib/1.2.5
22/09/2011 22:59:43 | | Data directory: D:\_TEMP\BOINC
22/09/2011 22:59:43 | | Running under account JamesR
22/09/2011 22:59:43 | | Processor: 8 GenuineIntel Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz [Family 6 Model 26 Stepping 5]
22/09/2011 22:59:43 | | Processor: 256.00 KB cache
22/09/2011 22:59:43 | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 syscall nx lm vmx tm2 popcnt pbe
22/09/2011 22:59:43 | | OS: Microsoft Windows 7: Ultimate x64 Edition, Service Pack 1, (06.01.7601.00)
22/09/2011 22:59:43 | | Memory: 11.99 GB physical, 23.98 GB virtual
22/09/2011 22:59:43 | | Disk: 1.82 TB total, 800.40 GB free
22/09/2011 22:59:43 | | Local time is UTC +1 hours
22/09/2011 22:59:43 | | NVIDIA GPU 0: GeForce GTX 260 (driver version 27550, CUDA version 4000, compute capability 1.3, 838MB, 605 GFLOPS peak)
22/09/2011 22:59:43 | | NVIDIA GPU 1 (not used): GeForce GTS 250 (driver version 27550, CUDA version 4000, compute capability 1.1, 967MB, 415 GFLOPS peak)
22/09/2011 22:59:43 | | Version change (6.12.34 -> 6.12.33)
22/09/2011 22:59:43 | GPUGRID | URL http://www.gpugrid.net/; Computer ID 107950; resource share 100
22/09/2011 22:59:43 | World Community Grid | URL http://www.worldcommunitygrid.org/; Computer ID 1307766; resource share 100
22/09/2011 22:59:43 | World Community Grid | General prefs: from World Community Grid (last modified 01-Jan-1970 00:00:01)
22/09/2011 22:59:43 | World Community Grid | Host location: none
22/09/2011 22:59:43 | World Community Grid | General prefs: using your defaults
22/09/2011 22:59:43 | | Reading preferences override file
22/09/2011 22:59:43 | | Preferences:
22/09/2011 22:59:43 | | max memory usage when active: 6139.54MB
22/09/2011 22:59:43 | | max memory usage when idle: 11051.17MB
22/09/2011 22:59:43 | | max disk usage: 40.00GB
22/09/2011 22:59:43 | | max CPUs used: 7
22/09/2011 22:59:43 | | suspend work if non-BOINC CPU load exceeds 25 %
22/09/2011 22:59:43 | | (to change preferences, visit the web site of an attached project, or select Preferences in the Manager)
22/09/2011 22:59:43 | | Not using a proxy
22/09/2011 22:59:44 | | Running CPU benchmarks
22/09/2011 22:59:44 | | Suspending computation - CPU benchmarks in progress
22/09/2011 23:00:15 | | Benchmark results:
22/09/2011 23:00:15 | | Number of CPUs: 7
22/09/2011 23:00:15 | | 2592 floating point MIPS (Whetstone) per CPU
22/09/2011 23:00:15 | | 8148 integer MIPS (Dhrystone) per CPU
22/09/2011 23:00:16 | World Community Grid | Started download of hcc1_image04_6.42.tga
22/09/2011 23:00:16 | World Community Grid | Started download of hcc1_image01_6.42.tga

This means that you have created a special disk file containing your BOINC preferences overriding your website preferences.

This means if if 2 threadas are running, you will suspend the 7 BOINC threads you are allowing. I hope you have selected 'Leave Application in Memory if Suspended' so your multi-GB virtual memory will save everything instead of dumping it back to the last checkpoint.

I think these are general data files downloaded before you run your first HCC work unit. Does your My Grid page show any time for the HCC project? What does your Results Status page show for WCG projects?


Lawrence

Just to be sure, did you *do/try all* of my last post in this thread?

--//--

Hi _Absinthe_, It is a perfectly fine method as long as you remember that you are doing it. I normally look askance at it because so many members 'set and forget'. The idea is that BOINC should 'just work' so people do not need to remember choices made years ago.

Something is going wrong with your network commo that only seems to affect WCG. I think that while you are trying to resolve it, you should allow BOINC to run continuously. To do this, allow BOINC to run unless non-BOINC programs exceed 90% (not 25%) of CPU capacity. At the same time, run only one BOINC project. If you have a hyperthreading quad-core, then this means no more than 13% (or 15% or 20%) of CPU capacity. This should keep things simple while you examine your network.

Check out your network BOINC preferences. Let them run wide open without limitations.

About saving BOINC check points - this is a little different than you think. A check point is not a mirror image copy of the computer state. It is a special point that a project algorithm reaches which allows a relatively small table or group of tables to save the current progress of the task. But this is highly dependent on the algorithm. Some projects reach this point several times a second or several times a minute. Saving at each point would involve a lot of unnecessary disk I/O, irritating many people. So BOINC allows them to set a minimum delay. The default is 1 minute. Who minds missing a minute of computation due to a power outage? After powering up, the project starts at the last saved check point.

On the other hand, some projects - such as CEP - might have an hour ( or 2 or 3 hours) between some check points. If members lose this much time just because the task was superseded by a higher priority email - well, some members complain. Which is why we have a 'Leave Application in Memory' option. With big Virtual Memory disk files, it seldom causes a problem.

So far, I have no sage advice to give you, so I am just advising you on ways to simplify your situation while you investigate.

Computer problems are so complicated and frustrating, it makes a lot of sense to keep everything we can control very simple while we try to figure it out.

Lawrence

oh, and I will be enabling some of the additional logging options and removing the 'fix' entry from my hosts file, to try and find the exact cause of the problem, but that'll have to wait til later as have work to do now :(