stdoutdae.txt

2007-04-03 02:38:22 [SETI@home] Computation for task 29ap04aa.25566.17361.129826.3.24_1 finished
2007-04-03 02:38:22 [---] Starting 29ap04aa.25566.17361.129826.3.156_1
2007-04-03 02:38:22 [SETI@home] Starting task 29ap04aa.25566.17361.129826.3.156_1 using setiathome_enhanced version 515
2007-04-03 02:38:24 [SETI@home] [file_xfer] Started upload of file 29ap04aa.25566.17361.129826.3.24_1_0
2007-04-03 02:38:31 [SETI@home] [file_xfer] Finished upload of file 29ap04aa.25566.17361.129826.3.24_1_0
2007-04-03 02:38:31 [SETI@home] [file_xfer] Throughput 5679 bytes/sec
2007-04-03 02:38:33 [SETI@home] Sending scheduler request: To report completed tasks
2007-04-03 02:38:33 [SETI@home] Reporting 1 tasks
2007-04-03 02:38:48 [SETI@home] Scheduler RPC succeeded [server version 509]
2007-04-03 02:38:48 [SETI@home] Deferring communication 11 sec, because requested by project
2007-04-03 02:39:01 [SETI@home] Computation for task 29ap04aa.25566.17361.129826.3.156_1 finished
2007-04-03 02:39:01 [---] Starting 29ap04aa.25566.17361.129826.3.150_3
2007-04-03 02:39:01 [SETI@home] Starting task 29ap04aa.25566.17361.129826.3.150_3 using setiathome_enhanced version 515
2007-04-03 02:39:03 [SETI@home] [file_xfer] Started upload of file 29ap04aa.25566.17361.129826.3.156_1_0
2007-04-03 02:39:08 [SETI@home] [file_xfer] Finished upload of file 29ap04aa.25566.17361.129826.3.156_1_0
2007-04-03 02:39:08 [SETI@home] [file_xfer] Throughput 9571 bytes/sec
2007-04-03 02:39:10 [SETI@home] Sending scheduler request: To report completed tasks
2007-04-03 02:39:10 [SETI@home] Reporting 1 tasks
2007-04-03 02:39:25 [SETI@home] Scheduler RPC succeeded [server version 509]
2007-04-03 02:39:25 [SETI@home] Deferring communication 11 sec, because requested by project
2007-04-03 02:55:03 [SETI@home] Computation for task 29ap04aa.25566.17361.129826.3.14_0 finished
2007-04-03 02:55:03 [---] Starting 29ap04aa.25566.17361.129826.3.141_2
2007-04-03 02:55:03 [SETI@home] Starting task 29ap04aa.25566.17361.129826.3.141_2 using setiathome_enhanced version 515
2007-04-03 02:55:06 [SETI@home] [file_xfer] Started upload of file 29ap04aa.25566.17361.129826.3.14_0_0

stderrdae.txt

Unhandled Exception Detected...

- Unhandled Exception Record -
Reason: Access Violation (0xc0000005) at address 0x0041221A read attempt to address 0x7069747C

Engaging BOINC Windows Runtime Debugger...



********************


BOINC Windows Runtime Debugger Version 5.9.0


Dump Timestamp    : 04/03/07 02:55:15

(I let run BOINC V5.9.0(.32 Crunch3rs) over V5.8.15)

The PC was running around one week without probs.
Then this morning I was looking to my PC and I saw that BOINC wasn't running.
(My firewall was asking me, I would like to give permission to let "Windows Help Center Service" to connect to internet)
Windows Task-Manager: computer 99% idle
Then I looked in the .txt-files and saw this...



The BOINC icon was in the taskbar but was frozen.
I couldn't click with the mouse on it.

Somebody know what was happened?

Thanx!

---

It might be because two units were finished very close to each other, and you are running crunch3r's ver of BOINC, with report immediately switched on.

See Error on file upload: Socket Read incomplete, but as you can see from Batman's posts etc you need to take this up with Crunch3r.

My advice would be go back to stock BOINC. Optimised benchmarks/crunch times are not needed on the most popular projects because of Fpops counting for credits, download qty is catered for with RDCF and report immediately only causes server overload.

Andy

It might be because two units were finished very close to each other, and you are running crunch3r's ver of BOINC, with report immediately switched on.

See Error on file upload: Socket Read incomplete, but as you can see from Batman's posts etc you need to take this up with Crunch3r.

My advice would be go back to stock BOINC. Optimised benchmarks/crunch times are not needed on the most popular projects because of Fpops counting for credits, download qty is catered for with RDCF and report immediately only causes server overload.

Andy

Hello!

The time between the 2 WUs is ~ 16 minutes...

So they are not so near.

(If I remember correct, I had WUs who finished in ~ 1 min. and were uploaded and reported immediately correct...)

You think this is ONLY a prob with "report-results-immediately"?
You think if I use the "normal" client, this prob with frozen BOINC will not come?


I have the QX6700, so I need "CPU-affinity"...
Or I can use the "normal" client too?
"Because the QX6700 is two times a Core2 Duo"

The clients who have "CPU-affinity" are only the BOINC V5.3.12.tx36 (Truxsoft) and the clients from Crunch3r.

I can delete the "report-results-immediately" in Trux´s client.
In Crunch3r´s not, or?
But I had a prob with Trux´s... I saw one time that he run only with 3 cores...
(But it was in the past maybe more often, because I cannot sit all the time in front of the PC... ;-)

Or I contact the server only one times a day... with network-access switched OFF/ON...

What could I do?


Friendly greetings!

---

Hi,
You only really need cpu affinity if you do more than one project on a 1:1:1:1 or 1:1 or maybe 100:50:50 share. It can be arranged with stock BOINC if you set your connect to time very low 0.05 days or even lower and network available always.
This way you normally have one/project processing and possible one/project either waiting to report or waiting to process.

I used to do this with a dual P3 and sons P4 HT on Seti and Einstein, but had to babysit after server outages to reset to one Seti and one Einstein running. It gave a boost to Seti, but not Einstein of about 10%. Found that with much larger L2 cache of C2D that gains from cpu affinity are much reduced and not really needed these days.

Reporting involves opening several databases at project HQ and therefore in is better from there point of view, and ours when servers are under pressure (recovering for outage or maintenance) to report as many as possible in one visit.
For this to happen you need a work cache and limited access to the network.

So the suggestions above conflict on their network access requirments.

From experience with my C2D I find that most units are reported within an hour, or less, of uploading because there is a request for new work.

I agree cpu affinity with multiple projects would be desirable, but what happens when there is a no work for a project. Does that cpu stop working or will it switch to the other project(s)?

The choices are with you my friend, you have to experiment, because most suggestions, including mine, will probably be countered by another person. In fact in the early days of BOINC, not many people had multiple cpu machines and didn't believe there could be a boost in performance by using cpu affinity especially when the cpu's were physically separate.

---

Hi,
You only really need cpu affinity if you do more than one project on a 1:1:1:1 or 1:1 or maybe 100:50:50 share. It can be arranged with stock BOINC if you set your connect to time very low 0.05 days or even lower and network available always.
This way you normally have one/project processing and possible one/project either waiting to report or waiting to process.

I used to do this with a dual P3 and sons P4 HT on Seti and Einstein, but had to babysit after server outages to reset to one Seti and one Einstein running. It gave a boost to Seti, but not Einstein of about 10%. Found that with much larger L2 cache of C2D that gains from cpu affinity are much reduced and not really needed these days.

Reporting involves opening several databases at project HQ and therefore in is better from there point of view, and ours when servers are under pressure (recovering for outage or maintenance) to report as many as possible in one visit.
For this to happen you need a work cache and limited access to the network.

So the suggestions above conflict on their network access requirments.

From experience with my C2D I find that most units are reported within an hour, or less, of uploading because there is a request for new work.

I agree cpu affinity with multiple projects would be desirable, but what happens when there is a no work for a project. Does that cpu stop working or will it switch to the other project(s)?

The choices are with you my friend, you have to experiment, because most suggestions, including mine, will probably be countered by another person. In fact in the early days of BOINC, not many people had multiple cpu machines and didn't believe there could be a boost in performance by using cpu affinity especially when the cpu's were physically separate.

Hello!

Thank you very much for your support!

I will think about to use the BOINC V5.8.15...

But before, I would like to know, whether it's possible to delete "report-results-immediately" in Crunch3r´s clients... ;-)
(to use the cpu-affinity :-)

So I will start a new thread about this...

"delete rep.res.immediately in Crunch3rs client?!"


Friendly greetings!

---

I was going to my PC and saw, that the S@H app was not running...
I looked to BOINC and saw the message (this time BOINC was not frozen):
"BOINC cannot connect to app"

I let run BOINC V5.3.12.tx36 BUT without "report-results-immediately", this function is OFF!

What was going wrong?


These are the last .txt- files- messages:


stdoutdae.txt

2007-04-09 21:33:08 [---] Rescheduling CPU: application exited
2007-04-09 21:33:08 [SETI@home] Computation for result XXXXXXXXXXXXXXXXXXXXXXXXX finished
2007-04-09 21:33:08 [SETI@home] Starting result YYYYYYYYYYYYYYYYYYYYYYYYY using setiathome_enhanced version 515
2007-04-09 21:33:11 [SETI@home] Started upload of XXXXXXXXXXXXXXXXXXXXXXXXX
2007-04-09 21:33:33 [---] Network error: couldn't connect to server
2007-04-09 21:33:33 [SETI@home] Temporarily failed upload of XXXXXXXXXXXXXXXXXXXXXXXXX: http error
2007-04-09 21:33:33 [SETI@home] Backing off 1 minutes and 0 seconds on upload of file XXXXXXXXXXXXXXXXXXXXXXXXX
2007-04-09 21:34:34 [SETI@home] Started upload of XXXXXXXXXXXXXXXXXXXXXXXXX
ERROR: file upload returned -127
2007-04-09 21:34:51 [SETI@home] Temporarily failed upload of XXXXXXXXXXXXXXXXXXXXXXXXX: transient upload error



stderrdae.txt

2007-04-09 21:33:33 [---] Network error: couldn't connect to server

***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x7C920F29 read attempt to address 0x645F6174

1: 04/09/07 21:34:51

---