Since 7/21, one machine A-SYS running 6.13.1 & Lunatics _x39e, keeps asking for CUDA work and the server says no joy. While the B-SYS running 6.12.33 & Lunatics _x38g is sucking up CUDA work like a vacumn cleaner on steroids. In the mean time the other machine is sucking up MB 6.03 & AP WUs with the same steroids. Any suggestions?

---

I don't buy computers, I build them!!

Luck of the draw? Out of my 20 some odd machines some get 0 tasks when they ask and other get work every time they ask. Just seems to be the state of the feeder at the exact moment they ask.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

why not reschedule units over to the GPU?

---

Since 7/21, one machine A-SYS running 6.13.1 & Lunatics _x39e, keeps asking for CUDA work and the server says no joy. While the B-SYS running 6.12.33 & Lunatics _x38g is sucking up CUDA work like a vacumn cleaner on steroids. In the mean time the other machine is sucking up MB 6.03 & AP WUs with the same steroids. Any suggestions?

Report your alpha testing observations for 6.13.1 to the developers. That totally new work fetch approach will obviously operate differently. So you send back the logs with the recommended log flags set, and the developers can analyze and fix it for 6.13.2 or later.

                                                                   Joe

Since 7/21, one machine A-SYS running 6.13.1 & Lunatics _x39e, keeps asking for CUDA work and the server says no joy. While the B-SYS running 6.12.33 & Lunatics _x38g is sucking up CUDA work like a vacumn cleaner on steroids. In the mean time the other machine is sucking up MB 6.03 & AP WUs with the same steroids. Any suggestions?

Report your alpha testing observations for 6.13.1 to the developers. That totally new work fetch approach will obviously operate differently. So you send back the logs with the recommended log flags set, and the developers can analyze and fix it for 6.13.2 or later.

                                                                   Joe

I don't think that the version of Lunatics is the problem, it has been running perfectly up until the 21st. I even uninstalled and regressed to v.037 and still nojoy, before going back to _x39e.

I just think the servers are being stingy with which machines it is sending CUDA to, but I would like it to make up its mind and send me some.

---

I don't buy computers, I build them!!

Since 7/21, one machine A-SYS running 6.13.1 & Lunatics _x39e, keeps asking for CUDA work and the server says no joy. While the B-SYS running 6.12.33 & Lunatics _x38g is sucking up CUDA work like a vacumn cleaner on steroids. In the mean time the other machine is sucking up MB 6.03 & AP WUs with the same steroids. Any suggestions?

Report your alpha testing observations for 6.13.1 to the developers. That totally new work fetch approach will obviously operate differently. So you send back the logs with the recommended log flags set, and the developers can analyze and fix it for 6.13.2 or later.

                                                                   Joe

I don't think that the version of Lunatics is the problem, it has been running perfectly up until the 21st. I even uninstalled and regressed to v.037 and still nojoy, before going back to _x39e.

I just think the servers are being stingy with which machines it is sending CUDA to, but I would like it to make up its mind and send me some.

Joe isn't commenting on your version of Lunatics, he's commenting on your version of BOINC. Read him again - he's right. Do that.

Since 7/21, one machine A-SYS running 6.13.1 & Lunatics _x39e, keeps asking for CUDA work and the server says no joy. While the B-SYS running 6.12.33 & Lunatics _x38g is sucking up CUDA work like a vacumn cleaner on steroids. In the mean time the other machine is sucking up MB 6.03 & AP WUs with the same steroids. Any suggestions?

Report your alpha testing observations for 6.13.1 to the developers. That totally new work fetch approach will obviously operate differently. So you send back the logs with the recommended log flags set, and the developers can analyze and fix it for 6.13.2 or later.

                                                                   Joe

I don't think that the version of Lunatics is the problem, it has been running perfectly up until the 21st. I even uninstalled and regressed to v.037 and still nojoy, before going back to _x39e.

I just think the servers are being stingy with which machines it is sending CUDA to, but I would like it to make up its mind and send me some.

Joe isn't commenting on your version of Lunatics, he's commenting on your version of BOINC. Read him again - he's right. Do that.

I apologize for any misunderstanding on Joe's post. Here are my settings for the cc_config.xml file, which one(s) do I turn on? Sched_ops has been turned on for several days, but I see no increase in the log because of it.


<cc_config>
<log_flags>
<task>0</task>
<file_xfer>0</file_xfer>
<sched_ops>1</sched_ops>
<coproc_debug>0</coproc_debug>
<cpu_sched>0</cpu_sched>
<cpu_sched_debug>0</cpu_sched_debug>
<dcf_debug>0</dcf_debug>
<sched_op_debug>0</sched_op_debug>
<state_debug>0</state_debug>
<http_debug>0</http_debug>
<http_xfer_debug>0</http_xfer_debug>
</log_flags>
<options>
<use_all_gpus>1</use_all_gpus>
</options>
</cc_config>

---

I don't buy computers, I build them!!

I apologize for any misunderstanding on Joe's post. Here are my settings for the cc_config.xml file, which one(s) do I turn on? Sched_ops has been turned on for several days, but I see no increase in the log because of it.

As in the Alpha Test Announcement