Howdy!

I've got 4 machines running CEP2 exclusively; all but one are uploading normally. There are currently 22 WUs on the troublesome box that are stuck trying to upload.

Messages tab in BOINC says:

2/28/2011 7:47:56 AM World Community Grid [error] Error reported by file upload server: [E201345_927_A.28.C22H12N2S2Se2.26.2.set1d06_1_4] error locking file

2/28/2011 7:47:56 AM World Community Grid Temporarily failed upload of E201345_927_A.28.C22H12N2S2Se2.26.2.set1d06_1_4: transient upload error

Sure hope this gits fixed quick - because the more these pile up, the more clogged my intertube's gonna be when they finally start uploading again.

Regards,

SMTB1963

... Calling Techs, Calling Techs ... - Thanks, Sek
My 2 stuck uploads had a side-effect ...

Before I read the details of the error message (server file lock), I took your advice and "upgraded" BOINC from 6.2.19 to 6.10.58. It didn't fix the stuck uploads.
After a while, I noticed that no new work was being downloaded, and I was out of CEP2 as the 2 stuck uploads were taking out 2 of the quota of 5. OK, there was an FAAH WU waiting to report that had taken 2 x the normal crunch time. I gradually increased the work cache in the preferences, but still no download action. No suspended tasks shown, new work allowed in Projects tab. I set <work_fetch_debug> in cc_config.xml. It did seem to be trying to fetch, but the diagnostics gave no clue to about the reason for failure. Something about no project to fetch.
I upgraded again to BOINC 6.2.19. At first it failed to start, with popup error messages. I re-ran the installer, selecting "Repair", and it's OK. It was trying to fetch work, and this time the diagnostics were much more helpful: they said there was a suspended task. None were visible, but then I selected each of the CEP2s that were shown as "Uploading". The Suspend/Resume button revealed that I had suspended one of these when trying to free the upload, and had forgotten to Resume it.

Put the 6.2.19 diagnostics back into BOINC 6.10.59+ and I'll be happy to try it again.
- Rick -

Funny, but yes in the Tasks view you can suspend a task in Upload state. I'm going to try this today with 6.12.14 alpha first generally suspending networking so as to be able to ''capture'' completed tasks and see what it says after suspending 1 and opening the network. Can't remember, but think the state of the left margin buttons is the only sign. Don't think there's neither a message log indication that work fetch is thusly disabled. Think that's a long overdue indicator needing addition.

--//--

added: And so I did on the 6.10.58 mounted Linux quad and suspended a task that was sitting in "uploading" state. The task view showing "Suspended" [edit: in the Tasks view status column] and the message log capturing...

570 WCG 28-02-2011 18:03 task ts01_a103_pr23a1_0 suspended by user

though when I opened the network again, the result files happily uploaded and the supposedly "suspended" task happily reporting per the below integral log from completion to acknowledgment:

565 WCG 28-02-2011 18:02 Computation for task ts01_a103_pr23a1_0 finished
566 WCG 28-02-2011 18:02 [dcf] DCF: 1.168194->1.175943, raw_ratio 1.245678, adj_ratio 1.066328
567 WCG 28-02-2011 18:02 Starting ts01_a104_pr89a0_1
568 WCG 28-02-2011 18:02 [cpu_sched] Starting ts01_a104_pr89a0_1 (initial)
569 WCG 28-02-2011 18:02 Starting task ts01_a104_pr89a0_1 using dddt2 version 617
570 WCG 28-02-2011 18:03 task ts01_a103_pr23a1_0 suspended by user
571 28-02-2011 18:06 Resuming network activity
572 WCG 28-02-2011 18:06 Started upload of ts01_a103_pr23a1_0_0
573 WCG 28-02-2011 18:06 Started upload of ts01_a103_pr23a1_0_1
574 WCG 28-02-2011 18:06 Finished upload of ts01_a103_pr23a1_0_1
575 WCG 28-02-2011 18:06 Started upload of ts01_a103_pr23a1_0_2
576 WCG 28-02-2011 18:06 Finished upload of ts01_a103_pr23a1_0_0
577 WCG 28-02-2011 18:06 Finished upload of ts01_a103_pr23a1_0_2
578 WCG 28-02-2011 18:06 [sched_op_debug] Starting scheduler request
579 WCG 28-02-2011 18:06 Sending scheduler request: To report completed tasks.
580 WCG 28-02-2011 18:06 Reporting 1 completed tasks, not requesting new tasks
581 WCG 28-02-2011 18:06 [sched_op_debug] CPU work request: 0.00 seconds; 0.00 CPUs
582 WCG 28-02-2011 18:06 Scheduler request completed
583 WCG 28-02-2011 18:06 [sched_op_debug] Server version 601
584 WCG 28-02-2011 18:06 Project requested delay of 11 seconds
585 WCG 28-02-2011 18:06 [sched_op_debug] handle_scheduler_reply(): got ack for result ts01_a103_pr23a1_0

Network is suspended on the 6.12.14 client but tasks wont finish for a few hours, so then will learn what it does for this build. 6.12 has this new "notices" tab and pop ups so that will be an area to test if it does what it says on the can.

Well did the same thing on the 6.12.14 client:

- Network Suspended
- Task completed showing in Tasks view as Uploading (Suspended Network) in the status column
- Selected task and hit Suspend button, nothing changes, but the state button indicating to "Resume" whilst the Uploading task is selected.
- Unsuspended Network
- Horses not held, upload started and reporting completed without any intervention.

Not sure what 6.2.19 does differently, but suspending an uploading task does nothing in 6.10.58 and 6.12.14 when the line is open, not here.

--//--

Ad: Integral message log:

28/02/2011 20:03:07 | World Community Grid | Computation for task ts01_a235_pr56a0_0 finished
28/02/2011 20:03:07 | World Community Grid | [dcf] DCF: 0.730806->1.130143, raw_ratio 1.130143, adj_ratio 1.546434
28/02/2011 20:03:07 | World Community Grid | Starting ts01_a239_pr34b1_0
28/02/2011 20:03:07 | World Community Grid | [cpu_sched] Starting ts01_a239_pr34b1_0 (initial)
28/02/2011 20:03:07 | World Community Grid | Starting task ts01_a239_pr34b1_0 using dddt2 version 617
28/02/2011 20:06:33 | World Community Grid | [checkpoint] result ts01_a235_pr56b0_0 checkpointed
28/02/2011 20:07:05 | World Community Grid | task ts01_a235_pr56a0_0 suspended by user
28/02/2011 20:07:33 | | Resuming network activity
28/02/2011 20:07:34 | World Community Grid | Started upload of ts01_a235_pr56a0_0_0
28/02/2011 20:07:34 | World Community Grid | Started upload of ts01_a235_pr56a0_0_1
28/02/2011 20:07:41 | World Community Grid | Finished upload of ts01_a235_pr56a0_0_1
28/02/2011 20:07:41 | World Community Grid | Started upload of ts01_a235_pr56a0_0_2
28/02/2011 20:07:45 | World Community Grid | Finished upload of ts01_a235_pr56a0_0_0
28/02/2011 20:07:46 | World Community Grid | Finished upload of ts01_a235_pr56a0_0_2
28/02/2011 20:07:50 | World Community Grid | [sched_op] Starting scheduler request
28/02/2011 20:07:50 | World Community Grid | Sending scheduler request: To report completed tasks.
28/02/2011 20:07:50 | World Community Grid | Reporting 1 completed tasks, not requesting new tasks
28/02/2011 20:07:50 | World Community Grid | [sched_op] CPU work request: 0.00 seconds; 0.00 CPUs
28/02/2011 20:07:53 | World Community Grid | Scheduler request completed
28/02/2011 20:07:53 | World Community Grid | [sched_op] Server version 601
28/02/2011 20:07:53 | World Community Grid | Project requested delay of 11 seconds
28/02/2011 20:07:53 | World Community Grid | [sched_op] handle_scheduler_reply(): got ack for task ts01_a235_pr56a0_0
28/02/2011 20:07:53 | World Community Grid | [sched_op] Deferring communication for 11 sec
28/02/2011 20:07:53 | World Community Grid | [sched_op] Reason: requested by project

I'm glad yours got unstuck, but the error is still happening. e.g. I just retried a stuck transfer and still got:
Error reported by file upload server: [E201356_501_A.31.C22H10N4O2S3.120.1.set1d06_0_4] error locking file