Seeing no one jumped to make a ticket, I made one. ;-)
[trac]#728[/trac]

Thank you. This addition really can increase overall BOINC effectiveness.

---

I thought so:

16/09/2008 02:44:22|SETI@home|Sending scheduler request: To report completed tasks
16/09/2008 02:44:22|SETI@home|Reporting 1 tasks
16/09/2008 02:44:27|SETI@home|Scheduler RPC succeeded [server version 603]
16/09/2008 02:44:27|SETI@home|Deferring communication for 11 sec
16/09/2008 02:44:27|SETI@home|Reason: requested by project
[b][color=red]16/09/2008 03:03:11|climateprediction.net|Deferring communication for 1 min 0 sec[/color][/b]
16/09/2008 03:03:11|climateprediction.net|Reason: Unrecoverable error for result hadcm3istd_0f6b_1920_160_05937811_2 (The device does not recognize the command. (0x16) - exit code 22 (0x16))
16/09/2008 03:03:11|climateprediction.net|Computation for task hadcm3istd_0f6b_1920_160_05937811_2 finished
16/09/2008 03:03:11|climateprediction.net|Output file hadcm3istd_0f6b_1920_160_05937811_2_15.zip for task hadcm3istd_0f6b_1920_160_05937811_2 absent
16/09/2008 03:03:11|climateprediction.net|Output file hadcm3istd_0f6b_1920_160_05937811_2_16.zip for task hadcm3istd_0f6b_1920_160_05937811_2 absent
16/09/2008 03:03:11|Einstein@Home|Starting h1_1083.20_S5R4__1006_S5R4a_0
16/09/2008 03:03:13|Einstein@Home|Starting task h1_1083.20_S5R4__1006_S5R4a_0 using einstein_S5R4 version 604
16/09/2008 03:03:13|SETI@home|Sending scheduler request: To fetch work

and so on.

Notice how the one-minute comms deferral is initiated by the client, long before we get into any project-requested backoffs. If we did that for success results, as well as errors, the server databases would run smaller and cleaner [task ends --> reduce DCF (sometimes) --> scheduler contact to fetch work (sometimes) --> could double with reporting finished work, if delay introduced].

EDIT

And in case anyone doesn't believe my point about DCF, I've just watched this on another box:

16/09/2008 11:40:42|SETI@home|Computation for task 28au08ab.26001.11115.3.8.14_2 finished
16/09/2008 11:40:42|SETI@home|Starting 28au08ab.4533.2526.4.8.12_1
16/09/2008 11:40:42|SETI@home|Starting task 28au08ab.4533.2526.4.8.12_1 using setiathome_enhanced version 528
16/09/2008 11:40:44|SETI@home|[file_xfer] Started upload of file 28au08ab.26001.11115.3.8.14_2_0
16/09/2008 11:41:07|SETI@home|[file_xfer] Finished upload of file 28au08ab.26001.11115.3.8.14_2_0
16/09/2008 11:41:07|SETI@home|[file_xfer] Throughput 1917 bytes/sec
16/09/2008 11:42:22|SETI@home|Computation for task 20au08ae.31360.1299.9.8.45_1 finished
16/09/2008 11:42:22|SETI@home|Starting 28au08ab.4533.2526.4.8.24_0
16/09/2008 11:42:22|SETI@home|Starting task 28au08ab.4533.2526.4.8.24_0 using setiathome_enhanced version 528
16/09/2008 11:42:24|SETI@home|[file_xfer] Started upload of file 20au08ae.31360.1299.9.8.45_1_0
[b][color=red]16/09/2008 11:42:25|SETI@home|Sending scheduler request: To fetch work[/color][/b]
16/09/2008 11:42:25|SETI@home|Requesting 23285 seconds of new work, and reporting 1 completed tasks
16/09/2008 11:42:30|SETI@home|[file_xfer] Finished upload of file 20au08ae.31360.1299.9.8.45_1_0
16/09/2008 11:42:30|SETI@home|[file_xfer] Throughput 10478 bytes/sec
16/09/2008 11:42:30|SETI@home|Scheduler RPC succeeded [server version 603]
16/09/2008 11:42:30|SETI@home|Deferring communication for 11 sec
16/09/2008 11:42:30|SETI@home|Reason: requested by project

That was directly attributable to

16/09/2008 05:25:12|SETI@home|Computation for task ap_14au08ac_B4_P1_00396_20080912_08575.wu_1 finished
16/09/2008 05:25:14|SETI@home|[file_xfer] Started upload of file ap_14au08ac_B4_P1_00396_20080912_08575.wu_1_0
16/09/2008 05:25:22|SETI@home|[file_xfer] Finished upload of file ap_14au08ac_B4_P1_00396_20080912_08575.wu_1_0
16/09/2008 05:25:22|SETI@home|[file_xfer] Throughput 9065 bytes/sec

- an Astropulse task (much higher DCF) finishing six hours earlier. That sawtooth wave generator in action.