Maybe that's why this topic is called "use at your own risk" ?

You think!

I've had to comment out a bunch more of the creditnew code since the validation logic is utterly stupid when a project doesn't use FLOPS since all estimates (and also some of the work fetch logic and validation logic) use FLOPS measured on the device compared to estimated FLOPS for the workunit. When a project uses IOPS instead of FLOPS, those estimates can be off by an order of magnitude. When that happens BOINC thinks you are cheating (an outlier) so it won't grant credit even if the result was valid. Yes, it was logging that it was valid, then granting 0.0 credit and then changing the state back to inconclusive. I guess logic works different on the West Coast than the rest of the planet, as I would expect that a valid result _should_ actually get credit. I commented out all the "outlier" code, so it should work now. Let me know if it doesn't.

Credit has been granted to as many valid yet inconclusive results as I could find. It was a royal PITA to find the issue, identify the improper logic (specially since it is in code that they don't want projects to change) , and then search the validator log files containing the 60k results returned per day to find the 20 (that's 20, not 20k) that came from CPUs on average per day so I could manually grant the credit to them. I could only go back a few weeks since I don't archive the log files and once someone else completed the workunit, it is no longer on the server and the only history stored is about the person whose data was valid.

05/06/2018 18:42:54 | collatz | Tasks for CPU are available, but your preferences are set to not accept them
05/06/2018 18:42:54 | collatz | Tasks for NVIDIA GPU are available, but your preferences are set to not accept them
05/06/2018 18:42:54 | collatz | Tasks for Intel GPU are available, but your preferences are set to not accept them
05/06/2018 18:42:54 | collatz | New computer location: work

Notice the work location, is for PC with ATI 7790 GPU.
The projects\boinc.thesonntags.com_collatz dir continue to stay empty.

Enable sched_op_debug to get more detailed info. You will get more detail about it requesting work which may explain why it isn't getting work.

Have you installed the AMD OpenCL drivers? The ones installed by Windows will most likely be missing the OpenCL drivers.

If you still can't figure it out, what is the Host ID? (I'm not going to waste time wading though hidden hosts and search the log files for every computer you own)

05/06/2018 18:42:54 | collatz | Tasks for CPU are available, but your preferences are set to not accept them
05/06/2018 18:42:54 | collatz | Tasks for NVIDIA GPU are available, but your preferences are set to not accept them
05/06/2018 18:42:54 | collatz | Tasks for Intel GPU are available, but your preferences are set to not accept them
05/06/2018 18:42:54 | collatz | New computer location: work

Notice the work location, is for PC with ATI 7790 GPU.
The projects\boinc.thesonntags.com_collatz dir continue to stay empty.

Enable sched_op_debug to get more detailed info. You will get more detail about it requesting work which may explain why it isn't getting work.

Have you installed the AMD OpenCL drivers? The ones installed by Windows will most likely be missing the OpenCL drivers.

If you still can't figure it out, what is the Host ID? (I'm not going to waste time wading though hidden hosts and search the log files for every computer you own)

Yes, I've the same setup on my host working fine before the project stop,
same AMD OpenCL drivers and MS Visual CC++ xxxx Redistributable packages.
is an update necessary?

My host ID is 832917

23/06/2018 16:43:29 | | Starting BOINC client version 7.10.2 for windows_x86_64
23/06/2018 16:43:29 | | log flags: file_xfer, sched_ops, task, sched_op_debug
23/06/2018 16:43:29 | | Libraries: libcurl/7.47.1 OpenSSL/1.0.2g zlib/1.2.8
23/06/2018 16:43:29 | | Data directory: C:\Program Files\BOINCdata

23/06/2018 16:43:29 | | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 7700 Series (driver version 2117.13 (VM), device version OpenCL 2.0 AMD-APP (2117.13), 1024MB, 1024MB available, 1926 GFLOPS peak)

23/06/2018 16:44:53 | | Fetching configuration file from https://boinc.thesonntags.com/collatz/get_project_config.php
23/06/2018 16:45:17 | collatz | sched RPC pending: Project initialization
23/06/2018 16:45:17 | collatz | [sched_op] Starting scheduler request
23/06/2018 16:45:17 | collatz | [sched_op] Fetching master file
23/06/2018 16:45:19 | collatz | [sched_op] Got master file; parsing
23/06/2018 16:45:19 | collatz | [sched_op] Found 1 scheduler URLs in master file
23/06/2018 16:45:19 | collatz | Master file download succeeded
23/06/2018 16:45:25 | collatz | sched RPC pending: Project initialization
23/06/2018 16:45:25 | collatz | [sched_op] Starting scheduler request
23/06/2018 16:45:25 | collatz | Sending scheduler request: Project initialization.
23/06/2018 16:45:25 | collatz | Requesting new tasks for CPU and AMD/ATI GPU
23/06/2018 16:45:25 | collatz | [sched_op] CPU work request: 1.00 seconds; 0.00 devices
23/06/2018 16:45:25 | collatz | [sched_op] AMD/ATI GPU work request: 1.00 seconds; 0.00 devices
23/06/2018 16:45:27 | collatz | Scheduler request completed: got 0 new tasks
23/06/2018 16:45:27 | collatz | [sched_op] Server version 711
23/06/2018 16:45:27 | collatz | No tasks sent
23/06/2018 16:45:27 | collatz | Tasks for CPU are available, but your preferences are set to not accept them
23/06/2018 16:45:27 | collatz | Tasks for NVIDIA GPU are available, but your preferences are set to not accept them
23/06/2018 16:45:27 | collatz | Tasks for Intel GPU are available, but your preferences are set to not accept them
23/06/2018 16:45:27 | collatz | Project requested delay of 121 seconds
23/06/2018 16:45:27 | collatz | New computer location: work
23/06/2018 16:45:27 | collatz | [sched_op] Deferring communication for 00:02:01
23/06/2018 16:45:27 | collatz | [sched_op] Reason: requested by project
23/06/2018 16:47:32 | collatz | [sched_op] Starting scheduler request
23/06/2018 16:47:32 | collatz | Sending scheduler request: To fetch work.
23/06/2018 16:47:32 | collatz | Requesting new tasks for AMD/ATI GPU
23/06/2018 16:47:32 | collatz | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
23/06/2018 16:47:32 | collatz | [sched_op] AMD/ATI GPU work request: 864000.00 seconds; 1.00 devices
23/06/2018 16:47:34 | collatz | Scheduler request completed: got 0 new tasks
23/06/2018 16:47:34 | collatz | [sched_op] Server version 711
23/06/2018 16:47:34 | collatz | Project requested delay of 121 seconds
23/06/2018 16:47:34 | collatz | [sched_op] Deferring communication for 00:02:01
23/06/2018 16:47:34 | collatz | [sched_op] Reason: requested by project

23/06/2018 16:51:19 | collatz | update requested by user
23/06/2018 16:51:21 | collatz | sched RPC pending: Requested by user
23/06/2018 16:51:21 | collatz | [sched_op] Starting scheduler request
23/06/2018 16:51:21 | collatz | Sending scheduler request: Requested by user.
23/06/2018 16:51:21 | collatz | Requesting new tasks for AMD/ATI GPU
23/06/2018 16:51:21 | collatz | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
23/06/2018 16:51:21 | collatz | [sched_op] AMD/ATI GPU work request: 864000.00 seconds; 1.00 devices
23/06/2018 16:51:23 | collatz | Scheduler request completed: got 0 new tasks
23/06/2018 16:51:23 | collatz | [sched_op] Server version 711
23/06/2018 16:51:23 | collatz | Project requested delay of 121 seconds
23/06/2018 16:51:23 | collatz | [sched_op] Deferring communication for 00:02:01
23/06/2018 16:51:23 | collatz | [sched_op] Reason: requested by project

I added the opencl_ati_gpu plan class specifications for both i686 and x64 versions for Linux just in case the ati_opencl was the cause. In theory, any project can make up any plan class they want. I'm not so sure that works in reality. I had the plan classes listed as opencl_amd and those weren't working for windows apps. So, let me know if that solves the issue. If not, send me a private message with the host id so I can set the BOINC scheduler to log debug information. Then after you do an update, I can check the server log.

But, I hope the plan class change will fix the issue. The previous server version I was using didn't use the plan_class_spec.xml file and I just coded the plan class info in C++. This is supposed to be easier and not require coding, but it sure seems likes it's more work! (That, or I code faster than I write valid XML).

You think!

Yep, I do! Thanks for being so patient (cough, cough). ;-)

I use Ubuntu 18.04 with Boinc 7.9.3.

I cant get WU for my Nvidia Gforce GTX1050:

Crunshing with Seti with GPU works fine:

My UserID is: 33848

Mo 25 Jun 2018 21:24:22 CEST | | Starting BOINC client version 7.9.3 for x86_64-pc-linux-gnu
Mo 25 Jun 2018 21:24:22 CEST | | log flags: file_xfer, sched_ops, task, sched_op_debug
Mo 25 Jun 2018 21:24:22 CEST | | Libraries: libcurl/7.58.0 OpenSSL/1.1.0g zlib/1.2.11 libidn2/2.0.4 libpsl/0.19.1 (+libidn2/2.0.4) nghttp2/1.30.0 librtmp/2.3
Mo 25 Jun 2018 21:24:22 CEST | | Data directory: /var/lib/boinc-client
Mo 25 Jun 2018 21:24:22 CEST | | CUDA: NVIDIA GPU 0: GeForce GTX 1050 (driver version 390.48, CUDA version 9.1, compute capability 6.1, 1997MB, 1698MB available, 1862 GFLOPS peak)
Mo 25 Jun 2018 21:24:22 CEST | | [libc detection] gathered: 2.27, Ubuntu GLIBC 2.27-3ubuntu1
Mo 25 Jun 2018 21:24:22 CEST | | Host name: XXX
Mo 25 Jun 2018 21:24:22 CEST | | Processor: 16 AuthenticAMD AMD Ryzen 7 1700 Eight-Core Processor [Family 23 Model 1 Stepping 1]
Mo 25 Jun 2018 21:24:22 CEST | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb hw_pstate sme ssbd vmmcall fsgsbase bmi1 avx2 smep bmi2 rdseed adx smap clflushopt sha_ni xsaveopt xsavec xgetbv1 xsaves clzero irperf xsaveerptr arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif overflow_recov succor smca
Mo 25 Jun 2018 21:24:22 CEST | | OS: Linux Ubuntu: Ubuntu 18.04 LTS [4.15.0-23-generic|libc 2.27 (Ubuntu GLIBC 2.27-3ubuntu1)]
Mo 25 Jun 2018 21:24:22 CEST | | Memory: 15.68 GB physical, 2.00 GB virtual
Mo 25 Jun 2018 21:24:22 CEST | | Disk: 915.40 GB total, 659.34 GB free
Mo 25 Jun 2018 21:24:22 CEST | | Local time is UTC +2 hours
Mo 25 Jun 2018 21:24:22 CEST | | Config: GUI RPCs allowed from:
Mo 25 Jun 2018 21:24:22 CEST | collatz | URL https://boinc.thesonntags.com/collatz/; Computer ID 834229; resource share 100

Mo 25 Jun 2018 21:26:30 CEST | collatz | update requested by user
Mo 25 Jun 2018 21:26:32 CEST | collatz | sched RPC pending: Requested by user
Mo 25 Jun 2018 21:26:32 CEST | collatz | [sched_op] Starting scheduler request
Mo 25 Jun 2018 21:26:32 CEST | collatz | Sending scheduler request: Requested by user.
Mo 25 Jun 2018 21:26:32 CEST | collatz | Requesting new tasks for NVIDIA GPU
Mo 25 Jun 2018 21:26:32 CEST | collatz | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
Mo 25 Jun 2018 21:26:32 CEST | collatz | [sched_op] NVIDIA GPU work request: 777600.00 seconds; 1.00 devices
Mo 25 Jun 2018 21:26:34 CEST | collatz | Scheduler request completed: got 0 new tasks
Mo 25 Jun 2018 21:26:34 CEST | collatz | [sched_op] Server version 711
Mo 25 Jun 2018 21:26:34 CEST | collatz | Project requested delay of 121 seconds
Mo 25 Jun 2018 21:26:34 CEST | collatz | [sched_op] Deferring communication for 00:02:01
Mo 25 Jun 2018 21:26:34 CEST | collatz | [sched_op] Reason: requested by project
Mo 25 Jun 2018 21:28:40 CEST | collatz | [sched_op] Starting scheduler request
Mo 25 Jun 2018 21:28:40 CEST | collatz | Sending scheduler request: To fetch work.
Mo 25 Jun 2018 21:28:40 CEST | collatz | Requesting new tasks for NVIDIA GPU
Mo 25 Jun 2018 21:28:40 CEST | collatz | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
Mo 25 Jun 2018 21:28:40 CEST | collatz | [sched_op] NVIDIA GPU work request: 777600.00 seconds; 1.00 devices
Mo 25 Jun 2018 21:28:42 CEST | collatz | Scheduler request completed: got 0 new tasks
Mo 25 Jun 2018 21:28:42 CEST | collatz | [sched_op] Server version 711
Mo 25 Jun 2018 21:28:42 CEST | collatz | Project requested delay of 121 seconds
Mo 25 Jun 2018 21:28:42 CEST | collatz | [sched_op] Deferring communication for 00:02:01
Mo 25 Jun 2018 21:28:42 CEST | collatz | [sched_op] Reason: requested by project

What can i do?

> CUDA: NVIDIA GPU 0: GeForce GTX 1050 (driver version 390.48, CUDA version 9.1, compute capability 6.1, 1997MB, 1698MB available, 1862 GFLOPS peak)

Collatz only has OpenCL apps, not CUDA and from the description above, there's no OpenCL installed. Check out https://wiki.tiker.net/OpenCLHowTo

There's no join in Mudville!

The previous fix for the CPU credits wasn't working. Or rather, it didn't fix (and by fix I mean remove the stupid BOINC code that assumes all projects use floating point arithmetic). So, I commented out several hundred more lines of creditnew madness in both the validator.cpp and credit.cpp BOINC source code and then recompiled the server daemons.

I also manually changed the credits for the 204 WUs that were valid but not granted any credit.

If you run into a a problem, please provide the host id and result id as it makes it a lot easier to track down the problems. (Thanks, Conan for providing that info which led to this latest fix.)

Validator and assimilator are offline.

Computing status
Work
Tasks ready to send 799
Tasks in progress 149608
***Workunits waiting for validation 14886***
Workunits waiting for assimilation 19
Workunits waiting for file deletion 0
Tasks waiting for file deletion 0
Transitioner backlog (hours) 0.00

Update:

Tasks ready to send 800
Tasks in progress 151204
**Workunits waiting for validation 19536**
Workunits waiting for assimilation 13
Workunits waiting for file deletion 0
Tasks waiting for file deletion 0
Transitioner backlog (hours) 0.00

Morning update:

Tasks ready to send 799
Tasks in progress 150896
****Workunits waiting for validation 42725****
Workunits waiting for assimilation 15

Slicker, it would seem that a server reboot might be needed to restart the validator and the assimilator

Both of them display a status of not running and have been in that state for 12 hours or so.

Tasks ready to send 770
Tasks in progress 150977
Workunits waiting for validation 44650
Workunits waiting for assimilation 16

Apparently, Slicker is gone for the 4th of July holiday. So we are all out of luck until he returns, Hope we don't lose any more credits!

---

Brent

Right -- that's my guess as well -- I restarted my Moo and GPUGrid streams in the interim.

The issue is that someone is trying to hack the output and their crap is causing the validator to crash. I'm trying to add code to filter their invalid results so it will stop crashing on every bad WU returned.

Your filtering efforts have seemed to have had the unfortunate side effect of invalidating all the my "pending validation" WUs :(. Here's hoping you're successful and the side effects are temporary :). Lovely way to spend the 4th, Jon. Again, I can't thank you enough for the effort and support you offer to the BOINC community!

---

Bummer, Seems like a lot of WU's that were ready to be validated lost to invalid. Going to do something else until I know this has been fixed...................

---

https://signature.statseb.fr/sig-872.png