If you set the CI to 0 then it reports immediately. Another of the things people wanted.

Andy

I would be surprised if this were true, as the project admins have always maintained that this was not the best thing as far as server load was concerned.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Computer is connected to the Internet about every 0 days
Maintain enough work for an additional 1 days (for example one day)

Then BOINC 'report results immediately'?

Yes if you use BOINC version 5.10.7

---

If you set the CI to 0 then it reports immediately. Another of the things people wanted.

Andy

I would be surprised if this were true, as the project admins have always maintained that this was not the best thing as far as server load was concerned.

It is definitely working that way on both my computers. As soon as the 11 sec standoff is over they report. And as you can see requests for work are not accompanied by reporting.

22/06/2007 09:32:28|SETI@home|Resuming task 10ap99ab.14928.19120.765914.3.85_0 using setiathome_enhanced version 517
22/06/2007 10:44:50|SETI@home|Computation for task 10ap99ab.14928.19120.765914.3.85_0 finished
22/06/2007 10:44:50|SETI@home|Starting 14ap99aa.15872.8338.273560.3.170_1
22/06/2007 10:44:50|SETI@home|Starting task 14ap99aa.15872.8338.273560.3.170_1 using setiathome_enhanced version 517
22/06/2007 10:44:52|SETI@home|[file_xfer] Started upload of file 10ap99ab.14928.19120.765914.3.85_0_0
22/06/2007 10:44:57|SETI@home|[file_xfer] Finished upload of file 10ap99ab.14928.19120.765914.3.85_0_0
22/06/2007 10:44:57|SETI@home|[file_xfer] Throughput 8071 bytes/sec
22/06/2007 10:45:06|SETI@home|Sending scheduler request: To report completed tasks
22/06/2007 10:45:06|SETI@home|Reporting 1 tasks
22/06/2007 10:45:11|SETI@home|Scheduler RPC succeeded [server version 509]
22/06/2007 10:45:11|SETI@home|Deferring communication for 11 sec
22/06/2007 10:45:11|SETI@home|Reason: requested by project
22/06/2007 10:47:26|SETI@home|Sending scheduler request: To fetch work
22/06/2007 10:47:26|SETI@home|Requesting 64 seconds of new work
22/06/2007 10:47:31|SETI@home|Scheduler RPC succeeded [server version 509]
22/06/2007 10:47:31|SETI@home|Deferring communication for 11 sec
22/06/2007 10:47:31|SETI@home|Reason: requested by project
22/06/2007 10:47:33|SETI@home|[file_xfer] Started download of file 10ap99ab.14928.22384.103400.3.164
22/06/2007 10:47:37|SETI@home|[file_xfer] Finished download of file 10ap99ab.14928.22384.103400.3.164
22/06/2007 10:47:37|SETI@home|[file_xfer] Throughput 155211 bytes/sec
22/06/2007 12:23:50|SETI@home|Sending scheduler request: To fetch work
22/06/2007 12:23:50|SETI@home|Requesting 43 seconds of new work
22/06/2007 12:23:55|SETI@home|Scheduler RPC succeeded [server version 509]
22/06/2007 12:23:55|SETI@home|Deferring communication for 11 sec
22/06/2007 12:23:55|SETI@home|Reason: requested by project
22/06/2007 12:23:58|SETI@home|[file_xfer] Started download of file 14ap99aa.15872.13714.629824.3.29
22/06/2007 12:24:01|SETI@home|[file_xfer] Finished download of file 14ap99aa.15872.13714.629824.3.29
22/06/2007 12:24:01|SETI@home|[file_xfer] Throughput 147986 bytes/sec
22/06/2007 12:47:09|SETI@home Beta Test|Resuming task 15fe07aa.24859.12751.5.5.246_2 using setiathome_enhanced version 521
22/06/2007 13:49:03|SETI@home Beta Test|Computation for task 15fe07aa.24859.12751.5.5.246_2 finished
22/06/2007 13:49:03|SETI@home|Resuming task 14ap99aa.15872.8338.273560.3.170_1 using setiathome_enhanced version 517
22/06/2007 13:49:05|SETI@home Beta Test|[file_xfer] Started upload of file 15fe07aa.24859.12751.5.5.246_2_0
22/06/2007 13:49:11|SETI@home Beta Test|[file_xfer] Finished upload of file 15fe07aa.24859.12751.5.5.246_2_0
22/06/2007 13:49:11|SETI@home Beta Test|[file_xfer] Throughput 15326 bytes/sec
22/06/2007 13:49:16|SETI@home Beta Test|Sending scheduler request: To report completed tasks
22/06/2007 13:49:16|SETI@home Beta Test|Reporting 1 tasks
22/06/2007 13:49:21|SETI@home Beta Test|Scheduler RPC succeeded [server version 509]
22/06/2007 13:49:21|SETI@home Beta Test|Deferring communication for 7 sec
22/06/2007 13:49:21|SETI@home Beta Test|Reason: requested by project
22/06/2007 14:05:51|SETI@home|Computation for task 14ap99aa.15872.8338.273560.3.170_1 finished
22/06/2007 14:05:51|SETI@home|Starting 19jl99ab.28690.2529.823572.3.255_2
22/06/2007 14:05:52|SETI@home|Starting task 19jl99ab.28690.2529.823572.3.255_2 using setiathome_enhanced version 517
22/06/2007 14:05:54|SETI@home|[file_xfer] Started upload of file 14ap99aa.15872.8338.273560.3.170_1_0
22/06/2007 14:06:00|SETI@home|[file_xfer] Finished upload of file 14ap99aa.15872.8338.273560.3.170_1_0
22/06/2007 14:06:00|SETI@home|[file_xfer] Throughput 6656 bytes/sec
22/06/2007 14:06:03|SETI@home|Sending scheduler request: To report completed tasks
22/06/2007 14:06:03|SETI@home|Reporting 1 tasks
22/06/2007 14:06:08|SETI@home|Scheduler RPC succeeded [server version 509]
22/06/2007 14:06:08|SETI@home|Deferring communication for 11 sec
22/06/2007 14:06:08|SETI@home|Reason: requested by project


Andy

edit] msg times are BST (UTC +1) [/edit

...
Why would you want an optimized Boinc client? It has nothing to do with the crunching, so speeding it up with optimization is a waste of developers time. A 64 bit client is available.

Now I use BOINC V5.8.11 with Crunch3rs V5.9.0

He enabled 'CPU-affinity' and 'report results immediately'.

My question because of 'optimized' was only to enable this features in the new BOINC V5.10.7 too.


BTW.
Because I have a Quad-Core..
The Benchmark-bug is in this new V5.10.7 repaired?

Well if you think about there is no need for the -report_results_immediately switch anymore now that you can decouple the cache size from the the CI.

With the old version you could in essence report immediately by setting the minimum CI of 0.01 days, but the drawback was you carried no cache in that case.

I haven't experimented with it yet, but I would guess that setting the CI to zero and the cache to what ever you want must have the same effect. The reason is that the CI sets the maximum interval BOINC will hold on to a result which is ready to report.

IOW, with the old version BOINC could theoretically hold on to a completed result for up to the CI before reporting it. In practice it works out that it will hold on to completed results until the next work fetch cycle for the project in question, for most reasonably fast hosts. So unless when you set the CI to zero it tells BOINC to not report completed work until the next work fetch cycle then it should report it as soon as it's done. If that's the case, then the workaround is to just set a short CI (it looks like it can be 0.0001 days now, if the pref override dialog is any indication).

Alinator

...
Why would you want an optimized Boinc client? It has nothing to do with the crunching, so speeding it up with optimization is a waste of developers time. A 64 bit client is available.

Now I use BOINC V5.8.11 with Crunch3rs V5.9.0

He enabled 'CPU-affinity' and 'report results immediately'.

My question because of 'optimized' was only to enable this features in the new BOINC V5.10.7 too.


BTW.
Because I have a Quad-Core..
The Benchmark-bug is in this new V5.10.7 repaired?

Well if you think about there is no need for the -report_results_immediately switch anymore now that you can decouple the cache size from the the CI.

With the old version you could in essence report immediately by setting the minimum CI of 0.01 days, but the drawback was you carried no cache in that case.

I haven't experimented with it yet, but I would guess that setting the CI to zero and the cache to what ever you want must have the same effect. The reason is that the CI sets the maximum interval BOINC will hold on to a result which is ready to report.

IOW, with the old version BOINC could theoretically hold on to a completed result for up to the CI before reporting it. In practice it works out that it will hold on to completed results until the next work fetch cycle for the project in question, for most reasonably fast hosts. So unless when you set the CI to zero it tells BOINC to not report completed work until the next work fetch cycle then it should report it as soon as it's done. If that's the case, then the workaround is to just set a short CI (it looks like it can be 0.0001 days now, if the pref override dialog is any indication).

Alinator

I'm running 5.10.7 with a CI of 0 (as recommended by prefs page because of always on broadband internet connection), with the cache additional setting set for 0.5 days, and 3 projects. The only time I have seen it NOT report results immediately is when that particular project is down and it can't report them successfully.

---

https://youtu.be/iY57ErBkFFE

#Texit

Don't blame me, I voted for Johnson(L) in 2016.

Truth is dangerous... especially when it challenges those in power.

Well there you go...

Thought exercises do work! :-)

<edit> I've been looking at that recommendation on the web pref page since it appeared, and thought to myself, "Well that pretty much explodes the dire end of the world as we know it effects of reporting results immediately myth. ;-)

Alinator

Computer is connected to the Internet about every 0 days
Maintain enough work for an additional 1 days (for example one day)

Then BOINC 'report results immediately'?

Yes if you use BOINC version 5.10.7

Actually, not to argue nor throw a monkey wrench into things; however, after reading this Thread's early Posts I changed my Preferences to Computer is "connected to the Internet about every 0 days
Maintain enough work for an additional 3 days". Then on BOINC 5.4.11 hit "Update", (on Farragut - I'm up North at my Parents' place), and the Dell system took to it. BOINC 5.4.11 has been reporting finished WUs immediately for almost 24 Hours. 6 WUs have reported immediately from the change made yesterday to present. I wanted to test it out to see what would happen; so, it seems to be working.

With the above stated, I have a question for anyone currently running 5.10.7. Are all of the 5.8.x bugs, (Merging issues, Message Timeouts, and all the other bugs), remedied on 5.10.7? I plan to upgrade to 5.10.7 soon; however, I wanted to make sure the "coast was clear" and that these aforementioned 5.8.x issues are truly resolved first.


Sincerely,

---

TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos

Computer is connected to the Internet about every 0 days
Maintain enough work for an additional 1 days (for example one day)

Then BOINC 'report results immediately'?

Yes if you use BOINC version 5.10.7

Actually, not to argue nor throw a monkey wrench into things; however, after reading this Thread's early Posts I changed my Preferences to Computer is "connected to the Internet about every 0 days
Maintain enough work for an additional 3 days". Then on BOINC 5.4.11 hit "Update", (on Farragut - I'm up North at my Parents' place), and the Dell system took to it. BOINC 5.4.11 has been reporting finished WUs immediately for almost 24 Hours. 6 WUs have reported immediately from the change made yesterday to present. I wanted to test it out to see what would happen; so, it seems to be working.

The Maintain enough work for an additional 3 days setting will only work with BOINC 5.10.x.

---

Computer is connected to the Internet about every 0 days
Maintain enough work for an additional 1 days (for example one day)

Then BOINC 'report results immediately'?

Yes if you use BOINC version 5.10.7

Actually, not to argue nor throw a monkey wrench into things; however, after reading this Thread's early Posts I changed my Preferences to Computer is "connected to the Internet about every 0 days
Maintain enough work for an additional 3 days". Then on BOINC 5.4.11 hit "Update", (on Farragut - I'm up North at my Parents' place), and the Dell system took to it. BOINC 5.4.11 has been reporting finished WUs immediately for almost 24 Hours. 6 WUs have reported immediately from the change made yesterday to present. I wanted to test it out to see what would happen; so, it seems to be working.

The Maintain enough work for an additional 3 days setting will only work with BOINC 5.10.x.

Thanks for that clarification. You are correct. Farragut did not Download any extra WUs... She only carries the two WUs crunching simultaneously - she's a Dual Core Dell.

As to 5.10.7; as asked in my original post, are all of the 5.8.x bugs, (Merging issues, Message Timeouts, and all the other bugs), remedied on 5.10.7? I plan to upgrade to 5.10.7 soon; however, I wanted to make sure the "coast was clear" and that these aforementioned 5.8.x issues are truly resolved first. Thanks in advance.


Sincerely,

---

TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos

As to 5.10.7; as asked in my original post, are all of the 5.8.x bugs, (Merging issues, Message Timeouts, and all the other bugs), remedied on 5.10.7? I plan to upgrade to 5.10.7 soon; however, I wanted to make sure the "coast was clear" and that these aforementioned 5.8.x issues are truly resolved first. Thanks in advance.


Sincerely,

Hey Timelord.

I'm running 5.10.7, and have not experienced any problems yet. Today makes day #2 of running it, benchmarks all look the same from previous 5.8.x versions. I have yet to have any problems. I guess my only suggestion would be to try it out, computers all act differently, so who knows, it may work just fine now for ya, or it might not. :-) Let us know.....

Jeremy :-)

---

Computer is connected to the Internet about every 0 days
Maintain enough work for an additional 1 days (for example one day)

Then BOINC 'report results immediately'?

Yes if you use BOINC version 5.10.7

Actually, not to argue nor throw a monkey wrench into things; however, after reading this Thread's early Posts I changed my Preferences to Computer is "connected to the Internet about every 0 days
Maintain enough work for an additional 3 days". Then on BOINC 5.4.11 hit "Update", (on Farragut - I'm up North at my Parents' place), and the Dell system took to it. BOINC 5.4.11 has been reporting finished WUs immediately for almost 24 Hours. 6 WUs have reported immediately from the change made yesterday to present. I wanted to test it out to see what would happen; so, it seems to be working.

The Maintain enough work for an additional 3 days setting will only work with BOINC 5.10.x.

Thanks for that clarification. You are correct. Farragut did not Download any extra WUs... She only carries the two WUs crunching simultaneously - she's a Dual Core Dell.

As to 5.10.7; as asked in my original post, are all of the 5.8.x bugs, (Merging issues, Message Timeouts, and all the other bugs), remedied on 5.10.7? I plan to upgrade to 5.10.7 soon; however, I wanted to make sure the "coast was clear" and that these aforementioned 5.8.x issues are truly resolved first. Thanks in advance.


Sincerely,

Hmmm...

That's interesting to know the older 5x CC's will respond to the zero CI. I hadn't felt a need to experiment with that since the 0.01 CI has served the purpose for my 'slugs'.

Out of curiosity, how does the fetch/report cycle work in that case? IOW, does it complete the current result(s), and then fetch a new one and report at the same time or what?

Alinator

Computer is connected to the Internet about every 0 days
Maintain enough work for an additional 1 days (for example one day)

Then BOINC 'report results immediately'?

Yes if you use BOINC version 5.10.7

Actually, not to argue nor throw a monkey wrench into things; however, after reading this Thread's early Posts I changed my Preferences to Computer is "connected to the Internet about every 0 days
Maintain enough work for an additional 3 days". Then on BOINC 5.4.11 hit "Update", (on Farragut - I'm up North at my Parents' place), and the Dell system took to it. BOINC 5.4.11 has been reporting finished WUs immediately for almost 24 Hours. 6 WUs have reported immediately from the change made yesterday to present. I wanted to test it out to see what would happen; so, it seems to be working.

The Maintain enough work for an additional 3 days setting will only work with BOINC 5.10.x.

Thanks for that clarification. You are correct. Farragut did not Download any extra WUs... She only carries the two WUs crunching simultaneously - she's a Dual Core Dell.

As to 5.10.7; as asked in my original post, are all of the 5.8.x bugs, (Merging issues, Message Timeouts, and all the other bugs), remedied on 5.10.7? I plan to upgrade to 5.10.7 soon; however, I wanted to make sure the "coast was clear" and that these aforementioned 5.8.x issues are truly resolved first. Thanks in advance.


Sincerely,

Hmmm...

That's interesting to know the older 5x CC's will respond to the zero CI. I hadn't felt a need to experiment with that since the 0.01 CI has served the purpose for my 'slugs'.

Out of curiosity, how does the fetch/report cycle work in that case? IOW, does it complete the current result(s), and then fetch a new one and report at the same time or what?

Alinator

Yes, Farragut completes her existing cache of WUs, (one per core), then Uploads them, then immediately Reports them. Finally, she Downloads new WUs. All 6 of the auto-reported WUs behaved in this manor.


Sincerely,

---

TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos

As to 5.10.7; as asked in my original post, are all of the 5.8.x bugs, (Merging issues, Message Timeouts, and all the other bugs), remedied on 5.10.7? I plan to upgrade to 5.10.7 soon; however, I wanted to make sure the "coast was clear" and that these aforementioned 5.8.x issues are truly resolved first. Thanks in advance.


Sincerely,

Hey Timelord.

I'm running 5.10.7, and have not experienced any problems yet. Today makes day #2 of running it, benchmarks all look the same from previous 5.8.x versions. I have yet to have any problems. I guess my only suggestion would be to try it out, computers all act differently, so who knows, it may work just fine now for ya, or it might not. :-) Let us know.....

Jeremy :-)

Cool, thanks. I will set Farragut to No New Tasks, then let the present cache of two WUs complete and Report in. Then I will upgrade Farragut to 5.10.7. If Farragut fares well with the Upgrade; then, when I return to SoCAL, I'll upgrade Excalibur. Now, this is the big "IF"; if Excalibur fares well, I will endeavor to upgrade Andromeda, (a K6-2, 350MHz), and Intrepid, (a P-III 600), to 5.10.7 as well.


Thanks again,

---

TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos

Cool, thanks. I will set Farragut to No New Tasks, then let the present cache of two WUs complete and Report in. Then I will upgrade Farragut to 5.10.7. If Farragut fares well with the Upgrade; then, when I return to SoCAL, I'll upgrade Excalibur. Now, this is the big "IF"; if Excalibur fares well, I will endeavor to upgrade Andromeda, (a K6-2, 350MHz), and Intrepid, (a P-III 600), to 5.10.7 as well.


Thanks again,

Well, you should have no trouble on your K6-2/350. I have 5.10.7 running fine on a K6-2/500, dual boots to 98SE and 2K.

It installed right over 5.8.15 no problemo.

Alinator

Well, you should have no trouble on your K6-2/350. I have 5.10.7 running fine on a K6-2/500, dual boots to 98SE and 2K.

It installed right over 5.8.15 no problemo.

Alinator

Works on my P60, 48Mram, win98se, installed over 5.10.6(which also worked fine).

Well, you should have no trouble on your K6-2/350. I have 5.10.7 running fine on a K6-2/500, dual boots to 98SE and 2K.

It installed right over 5.8.15 no problemo.

Alinator

Works on my P60, 48Mram, win98se, installed over 5.10.6(which also worked fine).

Sorting of the Report deadline column does not work correct on my win98se host.
But I do not know if the problem is Boinc or something missing on my host.

---

Well, you should have no trouble on your K6-2/350. I have 5.10.7 running fine on a K6-2/500, dual boots to 98SE and 2K.

It installed right over 5.8.15 no problemo.

Alinator

Works on my P60, 48Mram, win98se, installed over 5.10.6(which also worked fine).

Sorting of the Report deadline column does not work correct on my win98se host.
But I do not know if the problem is Boinc or something missing on my host.

Hmmm...

All the columns seem to sort OK on mine when in 9x, but I installed as a service in Win2k on a FAT32 partition and start it manually when in 9x. Maybe that had something to do with it.

Alinator

Well, you should have no trouble on your K6-2/350. I have 5.10.7 running fine on a K6-2/500, dual boots to 98SE and 2K.

It installed right over 5.8.15 no problemo.

Alinator

Works on my P60, 48Mram, win98se, installed over 5.10.6(which also worked fine).

Sorting of the Report deadline column does not work correct on my win98se host.
But I do not know if the problem is Boinc or something missing on my host.

Hmmm...

All the columns seem to sort OK on mine when in 9x, but I installed as a service in Win2k on a FAT32 partition and start it manually when in 9x. Maybe that had something to do with it.

Alinator

I have noticed that if you have the "Accessible view" enabled on my XP machine that the columns do not sort. If Accessible view is not enabled they sort fine.

Dave

---

SETI Classic Stats:
Workunits processed 28,419
Total CPU time 216,314 hours

Update:

Farragut is now on 5.10.7!!! 8-D She's crunching two WU's; however, no additional cache has been Downloaded even with the setting at 3 Days... Will this change? That is, when will I/am I to experience the 3 Day Cache of WU's?


Sincerely,

---

TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos

What's the issue with it auto-aborting WU's??? It auto-aborted 28 WU's on my C2D and 8 on my AMD. All WU's were due in July.

---