Messages

There are too many items in my gadget for it to be useful. For anyone running a boinc farm, IMHO, one would like to know if all the cpu's and gpu's are busy at a glance and the amount of the backlog (ready tasks).  The number of gpu's and cpu's would have to be obtained when BoincTasks starts up (and updated if a system reboots).  CPU tasks and GPU tasks would just be summed up over active tasks.  The gadget would show something like the following

System  | CPUs/Used  | GPUs/Used |
           |     /Ready   |     /Ready  |
-------- |------------|------------|
jys2x275      4/4/16           2/2/24
linux_i7        8/7/40           1/1/12

..etc..

For the 2nd line item above one cpu is not being used but there is sufficient backlog.   This would be colored yellow since a project like aqua might not start unless all the cpus were available to use.  if the backlog was under 8 then that line would be colored red since there was insufficient work for the system.

anyway, that is my wish.

[EDIT]
ok - i just discovered that if I set the time to display the gadget to 0 then I get to see the number of running tasks.  However, only about 1/2 the systems are displayed as shown below.



While the above image shows "jstateson2quad, Running:2" there are actually 5 resources being used.  One gpu (one task) and 4 cpus running aqua (1 task but 4 threads)

Wish:  To have the CPU and GPU time displayed as another option in your statistical graphs.  I read where you have as an add "A minigraph column for the RAC"  however, that is not the same as the CPU and GPU time that BoincTasks displays for each task.

The statistics graph would have the following form:

Vertical Axis would show the time in seconds spent on the result.
Horizontal axis to show the number of work units processed since BoincStats was started.
GPU and CPU time would be plotted as separate lines.  If your graph class supports a legund then I suggest the average run time and its standard deviation be displayed as the legund.  Results that errored out would not be included in the graph.

Since some projects take only minutes (milkway with ati) and others take weeks, you might want to use log scale when combining projects on one graph if your class supports it.

I just lost 300+ collatz tasks here when one gpu had a driver error and the other gpu was running ok.  The tasks all errored out within seconds and were refilled by the collatz queue withing minutes.  I only found out when I spotted 5 red highlighted rows in boinctasks.  Downloads occured 5 at a time.

I also have had this happen in gpugrid but that project has a limit of, as I recall, 12 tasks total in 24 hours.  It can occur with a video driver error or when remote console is inadvertanly brought up on a system running a gpu (vnc is ok, remote console stops the gpu)


The Wish:   That after 10 compute errors in a row with the same project and the same system, that the project is suspended and highlighted and something flash.  The 10 would be adjustable.

Since clicking on the computer tree (side panel) does nothing, yet, I was wishing that I could launch VNC or some remote access program of my choosing.  Can you add another column after "Wanted Tasks" such as "Launch Task"?



To configure, one would have to click on a browse button so as to browse to the desired app (vnc, ultra vnc, etc) and supply as the argument the ip address.  For ultra vnc all that is needed is "C:\Program Files\UltraVNC>vncviewer.exe 192.168.0.4" and the browser could show something editable like "C:\Program Files\UltraVNC\vncviewer.exe %s" where one could add any operands  as needed.  For example (the 192.168.0.4 above):



this would then bring up, for example,



Since ultra vnc can keep passwords and credentials one would not have to supply the vnc password on the launch line.  All that would be needed is the launch command line which would be in your computers.xml file such as
<launch_task>C:\Program Files\UltraVNC\vncviewer.exe %s</launch_task>

The 99% was the value reported by boinctasks.

There is a discussion of this over at collatz where the developer states that the value for "time" returned by the ATI task is really GPU time and not CPU time.  This generates the %99 figure.  IMHO they need to change their code to report CPU time which is about 2 seconds for the entire work unit.  Milkyway correctly returns CPU time which then makes the boinctasks cpu% calculation correct, again, IMHO.

OK - I read where you stated here quote "So 10% means the CPU ran 10% of it's time to feed the GPU"

Looking at the performance of an ATI 5850 on a Q6700, I see that collatz is a CPU hog and takes a full CPU.  As shown in the graph, the ati13ati collatz is highlighted in red (99.72) while it is working its way up and one that finished is shown at the top (99.52).  Note that the CUDA collatz "2.02 collatz" is NOT a cpu hog and used only 1.71% of a cpu.  Are you sure this is correct interpretation?



If so, then I should set the number of cpu's available on my quad from 4 to 3.  If I leave it at 4 then the the 4 QuantumFIRE alpha all take longer to finish.  However, this same system runs the ati13ati version of milkyway which uses less than %8 (7.42 as shown below).  This puts me in a bind because I dont need to set aside a cpu for milkyway.




Unless there is something I misunderstand, I am going to post over at collatz that their ati version uses almost a full cpu.  It does run extremely fast, and generates a lot of credits, but takes 4 minutes longer than a milkyway for about the same credits and burns a full cpu.


[EDIT} - I just checked the task manager and I do NOT see anything like %99 utilization for collatz.

.47 installed and ran just fine.  .48 installed but will not run unless I used MSVC debugger to continue past the breaks.

Vista 64 with all patches and updates, quad cpu.

This shows up on every execution



I then select vs2008 debug



then position the address register to the xor to allow a return to the caller.  I suspect that the XOR is used to return an OK (eventually) to the caller.  However, it has to be done twice.



have to do the above twice, didnt bother taking a snapshot the 2nd time.  ie: position the cursor at the XOR and then click on "Set Next Statement"



finally shows up and runs ok.

If I qut the program then the same problem occures the next time I run it.
I have DEP turned on for only essential Windows programs and services ie:  boinctasks64.exe is not guarded by DEP.  It is also 64 bit so cannot be under DEP in any event.


From Vista64 event logger

Log Name:      Application
Source:        Application Error
Date:          4/12/2010 11:04:28 PM
Event ID:      1000
Task Category: (100)
Level:         Error
Keywords:      Classic
User:          N/A
Computer:      jstateson2Quad
Description:
Faulting application boinctasks64.exe, version 0.4.9.0, time stamp 0x4bc25a82, faulting module boinctasks64.exe, version 0.4.9.0, time stamp 0x4bc25a82, exception code 0xc0000005, fault offset 0x0000000000092c3f, process id 0x1720, application start time 0x01cadabe69b7e020.