I just want to add that this is happening on my windows xp computer, only. The windows 10 machine is downloading WUs with no problems, so far.

And sometimes, more than one file gets stuck.

I have another "ghost" task on one of my hosts.

I too just noticed I had a task that timed out without a response on me. I went over the BOINC log and couldn't find its name. I did notice the following in the log for December 29th however (when the task was assigned to me):

29-Dec-2015 13:30:40 [GPUGRID] Sending scheduler request: To fetch work.
29-Dec-2015 13:30:40 [GPUGRID] Requesting new tasks for NVIDIA GPU
29-Dec-2015 13:35:47 [GPUGRID] Scheduler request failed: Timeout was reached
29-Dec-2015 13:35:47 [GPUGRID] Sending scheduler request: To fetch work.
29-Dec-2015 13:35:47 [GPUGRID] Requesting new tasks for NVIDIA GPU
29-Dec-2015 13:35:49 [GPUGRID] Scheduler request completed: got 0 new tasks
29-Dec-2015 13:35:49 [GPUGRID] No tasks sent
29-Dec-2015 13:35:49 [GPUGRID] No tasks are available for Long runs (8-12 hours on fastest card)
29-Dec-2015 13:35:49 [GPUGRID] Project has no tasks available
29-Dec-2015 13:35:51 [---] Project communication failed: attempting access to reference site
29-Dec-2015 13:35:52 [---] Internet access OK - project servers may be temporarily down.

So, it seems to me the request for new tasks did go through to the scheduler, but its response never reached my machine.

I am also having the download / upload problems mentioned in this thread. Files eventually do get down / up, but with several retries.

This is definitely a network problem on the GPUGRID side of the network - maybe a router close to the project servers has not had its DNS and / or routing tables refreshed?

I am wondering how this issue with phantom WU assignments is affecting WU availability and the overall computation progress, especially in this WU season of drought. Just imagine hosts requesting tasks, getting them without knowing it, and after some minutes requesting again. This issue does not need to happen many times to many users to make many tasks disappear...

---

Still getting these errors on most of my downloads. Eventually they come through.

Thu 07 Jan 2016 02:42:14 PM CST | GPUGRID | Temporarily failed download of e20s36_e16s2p1f382-GERARD_CXCL12_DIMPROTO1-0-pdb_file: transient HTTP error
Thu 07 Jan 2016 02:42:14 PM CST | GPUGRID | Backing off 00:02:24 on download of e20s36_e16s2p1f382-GERARD_CXCL12_DIMPROTO1-0-pdb_file
Thu 07 Jan 2016 02:42:15 PM CST | GPUGRID | Temporarily failed download of e20s36_e16s2p1f382-GERARD_CXCL12_DIMPROTO1-0-psf_file: transient HTTP error
Thu 07 Jan 2016 02:42:15 PM CST | GPUGRID | Backing off 00:02:44 on download of e20s36_e16s2p1f382-GERARD_CXCL12_DIMPROTO1-0-psf_file
Thu 07 Jan 2016 02:42:17 PM CST | | Internet access OK - project servers may be temporarily down.
Thu 07 Jan 2016 02:42:30 PM CST | | Project communication failed: attempting access to reference site
Thu 07 Jan 2016 02:42:30 PM CST | GPUGRID | Temporarily failed download of e20s3_e16s18p1f222-GERARD_CXCL12_DIMPROTO1-0-pdb_file: transient HTTP error
Thu 07 Jan 2016 02:42:30 PM CST | GPUGRID | Backing off 00:02:54 on download of e20s3_e16s18p1f222-GERARD_CXCL12_DIMPROTO1-0-pdb_file
Thu 07 Jan 2016 02:42:31 PM CST | | Internet access OK - project servers may be temporarily down.

Now that you mention it, I am too. There are several earlier entries, this is just the most recent. I never paid any attention to it before. Whether it is a big problem or not I have no idea.

i7-4790-PC

194	GPUGRID	1/7/2016 4:39:35 PM	Temporarily failed download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-coor_file: transient HTTP error	
195	GPUGRID	1/7/2016 4:39:35 PM	Backing off 00:03:10 on download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-coor_file	
196	GPUGRID	1/7/2016 4:39:35 PM	Started download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-par_file	
197	GPUGRID	1/7/2016 4:39:39 PM	Finished download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-par_file	
198	GPUGRID	1/7/2016 4:39:39 PM	Started download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-conf_file_enc	
199			1/7/2016 4:39:40 PM	Project communication failed: attempting access to reference site	
200	GPUGRID	1/7/2016 4:39:40 PM	Finished download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-conf_file_enc	
201	GPUGRID	1/7/2016 4:39:40 PM	Started download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-metainp_file	
202			1/7/2016 4:39:41 PM	Internet access OK - project servers may be temporarily down.	
203	GPUGRID	1/7/2016 4:39:41 PM	Finished download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-metainp_file	
204	GPUGRID	1/7/2016 4:39:41 PM	Started download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-hills_file	
205	GPUGRID	1/7/2016 4:39:42 PM	Finished download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-hills_file	
206	GPUGRID	1/7/2016 4:39:42 PM	Started download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-xsc_file	
207	GPUGRID	1/7/2016 4:39:43 PM	Finished download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-xsc_file	
208	GPUGRID	1/7/2016 4:39:43 PM	Started download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-prmtop_file	
209	GPUGRID	1/7/2016 4:39:44 PM	Finished download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-prmtop_file	
210	GPUGRID	1/7/2016 4:39:57 PM	Temporarily failed download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-psf_file: transient HTTP error	
211	GPUGRID	1/7/2016 4:39:57 PM	Backing off 00:02:26 on download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-psf_file	
212			1/7/2016 4:40:01 PM	Project communication failed: attempting access to reference site	
213			1/7/2016 4:40:02 PM	Internet access OK - project servers may be temporarily down.	
214	GPUGRID	1/7/2016 4:42:24 PM	Started download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-psf_file	
215	GPUGRID	1/7/2016 4:42:28 PM	Finished download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-psf_file	
216	GPUGRID	1/7/2016 4:42:46 PM	Started download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-coor_file	
217	GPUGRID	1/7/2016 4:42:51 PM	Finished download of e18s56_e12s22p1f460-GERARD_CXCL12_DIMPROTO3-0-coor_file	

Still getting these errors on most of my downloads. Eventually they come through.

Same here. After the initial download times out I hit "retry" from BoincTasks transfers tab and the download resumes and finishes. Been doing this for a couple of weeks.

I have lost 2 WUs while downloading on my windows xp machine:

https://www.gpugrid.net/result.php?resultid=14844544

https://www.gpugrid.net/result.php?resultid=14843167


The WUs were both GERARD_A2AR_luf6806.


Here is the event log:

1/13/2016 9:14:52 PM | GPUGRID | Requesting new tasks for NVIDIA GPU
1/13/2016 9:14:54 PM | GPUGRID | Scheduler request completed: got 1 new tasks
1/13/2016 9:14:56 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-LICENSE
1/13/2016 9:14:56 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-COPYRIGHT
1/13/2016 9:14:58 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-LICENSE: permanent HTTP error
1/13/2016 9:14:58 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-COPYRIGHT: permanent HTTP error
1/13/2016 9:14:58 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-coor_file
1/13/2016 9:14:58 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-vel_file
1/13/2016 9:14:59 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-coor_file: permanent HTTP error
1/13/2016 9:14:59 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-vel_file: permanent HTTP error
1/13/2016 9:14:59 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-idx_file
1/13/2016 9:14:59 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-pdb_file
1/13/2016 9:15:00 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-idx_file: permanent HTTP error
1/13/2016 9:15:00 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-pdb_file: permanent HTTP error
1/13/2016 9:15:00 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-psf_file
1/13/2016 9:15:00 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-par_file
1/13/2016 9:15:01 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-psf_file: permanent HTTP error
1/13/2016 9:15:01 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-par_file: permanent HTTP error
1/13/2016 9:15:01 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-conf_file_enc
1/13/2016 9:15:01 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-metainp_file
1/13/2016 9:15:02 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-conf_file_enc: permanent HTTP error
1/13/2016 9:15:02 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-metainp_file: permanent HTTP error
1/13/2016 9:15:02 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-hills_file
1/13/2016 9:15:02 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-xsc_file
1/13/2016 9:15:03 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-hills_file: permanent HTTP error
1/13/2016 9:15:03 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-xsc_file: permanent HTTP error
1/13/2016 9:15:03 PM | GPUGRID | Started download of e1s42_4-GERARD_A2AR_luf6806_b1-0-prmtop_file
1/13/2016 9:15:04 PM | GPUGRID | Giving up on download of e1s42_4-GERARD_A2AR_luf6806_b1-0-prmtop_file: permanent HTTP error
1/13/2016 9:16:04 PM | GPUGRID | Sending scheduler request: To report completed tasks.
1/13/2016 9:16:04 PM | GPUGRID | Reporting 1 completed tasks

I would like to report, that I have occasionally download problems until this date (individual files get stuck). This was not a concern, when there have not been many WUs around, but now when the pipeline is full, it is quite boring.

I would like to report, that I have occasionally download problems until this date (individual files get stuck). This was not a concern, when there have not been many WUs around, but now when the pipeline is full, it is quite boring.

Same here. I've asked about it several times but never got a reply. Hours wasted that could be used crunching.

Yes me too, stuck file usually downloads after a few hours before the task running is finished however 2 times recently it has been stuck for over 4 hours and this left GPU idle for a few hours. I hate that. Crunching computer is running using electricity, belching fire into our skies and no work is being done.

A single upload file also often gets stuck. Haven't lost bonus credit yet because of it but gone close a few times. Not good.

If I had a slower GPU this server malfunction would make me consider crunching another project where the download/upload server works properly. Perhaps even F@H.

recently i saw one article For All Portable issue problems.But after one week the content got changed to some game content...may be be that is because of my browser issue ..pls try read this article that gives exact solutions...also inform me about the issue i am facing....the link is http://bit.do/solveportableissues

Still having the same download issue. Recently brought my XP machine back to crunch here. It has the same issue. Downloads get stuck for hours. This is the only project of 7 that I'm currently running that does this and since it's On 2 different machines/OSs the problem is not on my end. PLEASE FIX THIS!

I can confirm this; it is not on my end! The problem has aroused when the project changed the network. It seems to me that the new network cannot cope with the size of data transferred from the server to the user and vice versa.

I had up-load problems before, but assumed this is caused by the ADSL contracted. But since the network change, it happens also when downloading files from the server.

Two comments:
First, another project is now happy with the spare GPU time.
Second, although it was lengthy discussed in another forum, because of this download problem, I suggest, the maximal WUs per GPU should be increased to three as the fastest cards get a better load with parallel crunching.

I routinely see delays of 10 to 20 minutes or so on downloads and a few uploads. I see it on both wired and wireless connections. It is annoying when I am running my GTX 750 Tis and am trying to make the 24 hour limit.

Maybe their servers are just overloaded?

I've forwarded your complaints to our IT service. Indeed delays in download/upload could be caused by the new network. I'll keep you updated!

I've forwarded your complaints to our IT service. Indeed delays in download/upload could be caused by the new network. I'll keep you updated!

Thanks.

FYI: I'm still having network access problems:
1. Stalled up/downloads
2. There are 3 tasks in progress according to my host's status page, but in reality there's only two (this host has only one GPU)
These problems were severe on 9th Dec, but it seems that the reason of these problems are not related to DNS replication.

I am have been having the same problems since the change to the new network. But for me, it happens occasionally during downloads only (definitely more often than before), and it involves only one file in the WU downloading. To continue with the download, I can either press "renter now " when the status is "Download reentry in xx:xx:xx" or exit boinc and then run it again. Sometimes, I have to do this more than once. But most of the time, the downloads go smoothly.

I haven't lost a WU yet, due to this problem. Knock wood!

I just want to add that this is happening on my windows xp computer, only. The windows 10 machine is downloading WUs with no problems, so far.

And sometimes, more than one file gets stuck.

WU file downloading problem is now happening on both my windows xp and 10 computers, occasionally. See log:


2/21/2016 6:15:03 PM | GPUGRID | Finished download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-par_file
2/21/2016 6:15:03 PM | GPUGRID | Started download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-conf_file_enc
2/21/2016 6:15:04 PM | GPUGRID | Finished download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-conf_file_enc
2/21/2016 6:15:04 PM | GPUGRID | Started download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-metainp_file
2/21/2016 6:15:05 PM | GPUGRID | Finished download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-metainp_file
2/21/2016 6:15:05 PM | GPUGRID | Started download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-hills_file
2/21/2016 6:15:06 PM | GPUGRID | Finished download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-hills_file
2/21/2016 6:15:06 PM | GPUGRID | Started download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-xsc_file
2/21/2016 6:15:07 PM | GPUGRID | Finished download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-xsc_file
2/21/2016 6:15:07 PM | GPUGRID | Started download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-prmtop_file
2/21/2016 6:15:08 PM | GPUGRID | Finished download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-prmtop_file
2/21/2016 6:20:02 PM | GPUGRID | Temporarily failed download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file: transient HTTP error
2/21/2016 6:20:02 PM | GPUGRID | Backing off 00:02:40 on download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file
2/21/2016 6:20:03 PM | | Project communication failed: attempting access to reference site
2/21/2016 6:20:04 PM | | Internet access OK - project servers may be temporarily down.
2/21/2016 6:22:42 PM | GPUGRID | Started download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file
2/21/2016 6:27:55 PM | | Project communication failed: attempting access to reference site
2/21/2016 6:27:55 PM | GPUGRID | Temporarily failed download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file: transient HTTP error
2/21/2016 6:27:55 PM | GPUGRID | Backing off 00:04:30 on download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file
2/21/2016 6:27:56 PM | | Internet access OK - project servers may be temporarily down.
2/21/2016 6:28:51 PM | GPUGRID | Started download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file
2/21/2016 6:29:06 PM | GPUGRID | Temporarily failed download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file: transient HTTP error
2/21/2016 6:29:06 PM | GPUGRID | Backing off 00:13:42 on download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file
2/21/2016 6:29:07 PM | | Project communication failed: attempting access to reference site
2/21/2016 6:29:08 PM | | BOINC can't access Internet - check network connection or proxy configuration.
2/21/2016 6:29:18 PM | GPUGRID | Started download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file
2/21/2016 6:29:35 PM | GPUGRID | Finished download of e18s17_e11s8p1f343-GERARD_CXCL12_CHLCPUBCHEM_chalcone121-0-psf_file

Another trick to get the download to restart is to disconnect and reconnect the network internet connection, and then in the boinc manager under the transfer tab press the renter now button with the stalled file highlighted. Of course, you can wait for it to restart on its own, this merely speeds up the download.


This was not happening in such frequency before the network upgrade.

7 tries and 5 1/2 hours wasted trying to download 1 file because every time the download fails the wait period for the next try gets longer. This is ridiculous. I thinks it's time to move somewhere else until this issue is resolved. 2+ months is long enough for me.

7 tries and 5 1/2 hours wasted trying to download 1 file because every time the download fails the wait period for the next try gets longer. This is ridiculous. I thinks it's time to move somewhere else until this issue is resolved. 2+ months is long enough for me.

Our complaints were forwarded to the IT service a week ago, however this problem exists since the changes in the network. I guess it's a misconfigured routing table (or more of them), which is quite hard to spot, especially when not all traffic is affected by it. A spare project could help to reduce the idle GPU time, so when the network issues will be fixed at GPUGrid's campus, your host will automatically stop downloading from the other (spare, 0 resource share) project.

7 tries and 5 1/2 hours wasted trying to download 1 file because every time the download fails the wait period for the next try gets longer. This is ridiculous. I thinks it's time to move somewhere else until this issue is resolved. 2+ months is long enough for me.

Our complaints were forwarded to the IT service a week ago, however this problem exists since the changes in the network. I guess it's a misconfigured routing table (or more of them), which is quite hard to spot, especially when not all traffic is affected by it. A spare project could help to reduce the idle GPU time, so when the network issues will be fixed at GPUGrid's campus, your host will automatically stop downloading from the other (spare, 0 resource share) project.

Something else blew up last night. I awoke this morning to find 4 tasks ready to report and no new tasks running. That had to be at least 12+ dead hours of no crunching. The projects tab showed the next update would not be for 12 more hours. After doing a manual update the 4 tasks reported and new tasks were requested. Below is a partial copy of the messages:

672710 GPUGRID 2/24/2016 7:24:35 AM update requested by user
672711 GPUGRID 2/24/2016 7:24:40 AM Fetching scheduler list
672712 GPUGRID 2/24/2016 7:24:43 AM Master file download succeeded
672713 GPUGRID 2/24/2016 7:24:48 AM Sending scheduler request: Requested by user.
672714 GPUGRID 2/24/2016 7:24:48 AM Reporting 4 completed tasks
672715 GPUGRID 2/24/2016 7:24:48 AM Requesting new tasks for NVIDIA GPU
672716 GPUGRID 2/24/2016 7:24:50 AM Scheduler request completed: got 1 new tasks

What would cause the master file to be needed again? I'm assuming that was a/the reason for the 12 hour delay. New tasks were received but there are 7 files stuck again. *bangs head on desk*

Also I don't think using a 0 share standby will work because if I remember correctly BOINC will not allow new tasks from another project to download if it detects stuck downloads from the higher priority project.

FWIW if the current IT service can't get this resolved after 2 months maybe GPUGrid might consider switching to another service provider.

What would cause the master file to be needed again?

Ten consecutive failures to contact the scheduler. Note that's the request work/report work contact attempt, not the file download attempts this thread has mainly been about.

Check the full log in stdoutdae.txt - see when the problem started/ended. Unless you've suppressed it, BOINC will try to contact a 'neutral' web host (google.com) after each failure: if google is OK but gpugrid fails, then the project server is the suspect. But if google fails as well, then your own network connection and ISP may be at fault.

To test a little theory of mine - what OS is having these problems? Linux, Windows, OS X? Or all three? I'm Windows, and I see the downloads stalling sometimes - but the work is usually fully downloaded by the time I need it.

[Edit - OK, we don't support OS X here. Forget that one.]