Mikey,
The host is returning large numbers of validated tasks so how does that suggest there's a problem with "gpu driver version"??
How could it even be remotely possible for the driver to interfere with work fetch??
Bein unfamiliar with how the intricacies of Boinc works but knowing that some drivers are not the 'tested' ones I was grasping at straws. Also since his pc's are hidden there's no way for me to know, other than what's been posted, if this pc is reporting mostly valid tasks or a mix of valid and invalid tasks or whatever. You as a Moderator have that advantage the rest of us don't.
...my theory is that it's a bug in the client work fetch logic. and likely centers around all the modifications that were made to work around the work fetch bug when using the project_max_concurrent and max_concurrent statements in app_config.
I tend to agree but had discounted the max_concurrent modifications because the client is listed as 7.16.6. I don't know for sure but doesn't that version pre-date those modifications?
I'm now wondering if it might not be a standard 7.16.6 build but perhaps one with some additional performance tweaks that might be mis-behaving?
I don't follow what is happening in the 'BOINC Pentathlon' world so wasn't even aware there was one going on. Yep, I know, I'm living under a rock :-). However that certainly adds additional context as to what might be causing the problem.
I dont think the current bug is directly related to the usage (or lack thereof) of the project_max_concurrent or max_concurrent statements. skillz has said to me offline that he's not using either of these.
I think a new bug was created in fixing the previous bug. they fixed one thing, and broke something else in the process.
...my theory is that it's a bug in the client work fetch logic. and likely centers around all the modifications that were made to work around the work fetch bug when using the project_max_concurrent and max_concurrent statements in app_config.
I tend to agree but had discounted the max_concurrent modifications because the client is listed as 7.16.6. I don't know for sure but doesn't that version pre-date those modifications?
I'm now wondering if it might not be a standard 7.16.6 build but perhaps one with some additional performance tweaks that might be mis-behaving?
I don't follow what is happening in the 'BOINC Pentathlon' world so wasn't even aware there was one going on. Yep, I know, I'm living under a rock :-). However that certainly adds additional context as to what might be causing the problem.
I dont think the current bug is directly related to the usage (or lack thereof) of the project_max_concurrent or max_concurrent statements. skillz has said to me offline that he's not using either of these.
I think a new bug was created in fixing the previous bug. they fixed one thing, and broke something else in the process.
That's very possible with part-time programmers, things are forgotten and overlooked because it works 'right now'. Don't get me wrong I think very highly of the Boinc Programmers but think they need to be full-time to keep things updated and to address 'bugs' like this quickly.
... since his pc's are hidden there's no way for me to know, other than what's been posted,
So you didn't read his opening post?? Seems like 1077 valid to 6 invalid suggests no problem with the OpenCL libraries. When hosts are 'hidden', I can see exactly what you can see - no more, no less.
It's great that you try to help people solve problems. You just need to read what's been posted before jumping to wrong conclusions. An experienced person will ignore bad advice but an inexperienced person might take unnecessary actions - just wasting time and risking further damage or confusion.
mikey wrote:
You as a Moderator have that advantage the rest of us don't.
The only things I can do that you can't are:-
Read and approve/reject new or updated profiles that users submit.
Ban people who are persistently/maliciously breaking the 'rules'.
Remove from view any posts that contain abusive/offensive content.
Force myself to try to keep up-to-date with what gets posted so I can attempt to implement the above.
That's it! So how does that give me any "advantage"?
... since his pc's are hidden there's no way for me to know, other than what's been posted,
So you didn't read his opening post?? Seems like 1077 valid to 6 invalid suggests no problem with the OpenCL libraries. When hosts are 'hidden', I can see exactly what you can see - no more, no less.
It's great that you try to help people solve problems. You just need to read what's been posted before jumping to wrong conclusions. An experienced person will ignore bad advice but an inexperienced person might take unnecessary actions - just wasting time and risking further damage or confusion.
mikey wrote:
You as a Moderator have that advantage the rest of us don't.
The only things I can do that you can't are:-
Read and approve/reject new or updated profiles that users submit.
Ban people who are persistently/maliciously breaking the 'rules'.
Remove from view any posts that contain abusive/offensive content.
Force myself to try to keep up-to-date with what gets posted so I can attempt to implement the above.
That's it! So how does that give me any "advantage"?
My mistake I thought you could 'see' his pc's even thought they are hidden from the rest of us, thereby giving you the advantage of being able to diagnose problems or see how things are going.
Gary Roberts wrote:mikey
)
Bein unfamiliar with how the intricacies of Boinc works but knowing that some drivers are not the 'tested' ones I was grasping at straws. Also since his pc's are hidden there's no way for me to know, other than what's been posted, if this pc is reporting mostly valid tasks or a mix of valid and invalid tasks or whatever. You as a Moderator have that advantage the rest of us don't.
Gary Roberts
)
I dont think the current bug is directly related to the usage (or lack thereof) of the project_max_concurrent or max_concurrent statements. skillz has said to me offline that he's not using either of these.
I think a new bug was created in fixing the previous bug. they fixed one thing, and broke something else in the process.
_________________________________________________________________________
Ian&Steve C. wrote: Gary
)
That's very possible with part-time programmers, things are forgotten and overlooked because it works 'right now'. Don't get me wrong I think very highly of the Boinc Programmers but think they need to be full-time to keep things updated and to address 'bugs' like this quickly.
mikey wrote:... since his
)
So you didn't read his opening post?? Seems like 1077 valid to 6 invalid suggests no problem with the OpenCL libraries. When hosts are 'hidden', I can see exactly what you can see - no more, no less.
It's great that you try to help people solve problems. You just need to read what's been posted before jumping to wrong conclusions. An experienced person will ignore bad advice but an inexperienced person might take unnecessary actions - just wasting time and risking further damage or confusion.
The only things I can do that you can't are:-
That's it! So how does that give me any "advantage"?
Cheers,
Gary.
Gary Roberts wrote: mikey
)
My mistake I thought you could 'see' his pc's even thought they are hidden from the rest of us, thereby giving you the advantage of being able to diagnose problems or see how things are going.