Problem uploading work units for climate prediction. They upload from my computer to 100% but never go up to the server. They stay stuck in my computer in upload. One stayed there for 3 days consuming all my network resources. Now I have another stuck there I suspended it for now. I am not downloading any more work units until this problem gets fixed. Older versions work fine...just the new version has the problem.

My guess is that the firewall on your PC doesn't know about the new version, and is blocking traffic.

Which firewall do you use? What permissions has boinc.exe been given?

Do you use a proxy server? (most commonly 'yes' if you connect via a university or work PC).

---

Critical missing information: operating system and what projects you're having the problem with.

---

Mike,

This is a home computer, not work. The problem is happening on 2 different computers with the same version of boinc. Boinc has been given full access through the firewall. Have a work unit out there now that is complete...but I can't get it to upload. I'm not a new user...have been running this program for many years. Thanks for looking at the problem but I think it is something else.

Problem uploading work units for climate prediction. They upload from my computer to 100% but never go up to the server. They stay stuck in my computer in upload. One stayed there for 3 days consuming all my network resources. Now I have another stuck there I suspended it for now. I am not downloading any more work units until this problem gets fixed. Older versions work fine...just the new version has the problem.

How are these two computers connected to the net?

If plugged into a router, have you tried rebooting the router?

---

Someone has just reported a similar problem caused by Kaspersky antivirus:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=5949&nowrap=true#32087

Which antivirus do you use?

---

I have Norton 360. Rosetta and Einstein work units go in and out just fine on the same computers...so I don't think that is the problem. The error reads a problem converting the files to cdnet format. Is that something on my end or is it in the upload portion of the file itself or possibly the boinc manager? I have reset the router several times and opened the firewall. At this point I am without a clue...and now it is happening with 5.10.28 work units as well. Maybe it is something with my network service provider since no one is having this problem but me.

The error "a problem converting the files to cdnet format" occurs during the post-processing phase, and shouldn't affect uploads or downloads in any way. Where are you seeing this message?

Norton is a known problem due to exclusive locks on files it is scanning - although not one I've seen associated with uploads or downloads. Try adding the Boinc directories to both of Norton's exclusion lists.

---

The errors were listed under my failed workunits. I don't know what to do with the workunits that are sitting in transfers for days now. I aborted another failed unit yesterday and still have 3 more files waiting to upload.

The error "a problem converting the files to cdnet format" occurs during the post-processing phase, and shouldn't affect uploads or downloads in any way. Where are you seeing this message?

Norton is a known problem due to exclusive locks on files it is scanning - although not one I've seen associated with uploads or downloads. Try adding the Boinc directories to both of Norton's exclusion lists.

---

Sharlee, could you tell us your CPDN computer ID number please? This would allow us to look at your CPDN web pages with your models, trickles etc. You'll find it in the BOINC manager messages on line 7 or shortly after that. (This isn't confidential info and it's safe to post it here.)

If you have Windows XP, you could enable its own firewall and open up its firewall info window to check that it says running, then disable your Norton firewall for a few minutes to see whether the zip files upload. You'd also need to click Update for CPDN in the BOINC manager Projects tab.

If you have Vista I expect it also has a built-in firewall that you could enable temporarily.

In your BOINC manager messages are you seeing 'HTTP error' when you try to upload? (If this were the case I expect you'd have told us, but we need to be sure.)

-197 (0xffffff3b)is the error message for 2 of the failed work units. I am having a problem with 3 computers. 798233, 707181 are 2 of them. CPDN monitor error quit connection. How do I fix this?

Sharlee, could you tell us your CPDN computer ID number please? This would allow us to look at your CPDN web pages with your models, trickles etc. You'll find it in the BOINC manager messages on line 7 or shortly after that. (This isn't confidential info and it's safe to post it here.)

If you have Windows XP, you could enable its own firewall and open up its firewall info window to check that it says running, then disable your Norton firewall for a few minutes to see whether the zip files upload. You'd also need to click Update for CPDN in the BOINC manager Projects tab.

If you have Vista I expect it also has a built-in firewall that you could enable temporarily.

In your BOINC manager messages are you seeing 'HTTP error' when you try to upload? (If this were the case I expect you'd have told us, but we need to be sure.)

-197 (0xffffff3b)is the error message for 2 of the failed work units. I am having a problem with 3 computers. 798233, 707181 are 2 of them. CPDN monitor error quit connection. How do I fix this?

Strange. -197 is ERR_ABORTED_VIA_GUI.

---

...
Strange. -197 is ERR_ABORTED_VIA_GUI.

Note that Sharlee says:

... I aborted another failed unit yesterday and still have 3 more files waiting to upload.

Sharlee, note that if you abort a transfer, you'll abort the associated workunit. The problem lies with network communication somewhere, i.e., firewall, A/V or proxy server. As far as I know Rosetta and Einstein don't transfer files via trickle_ups, so if the firewall, proxy or AV is blocking trickles they won't be affected.

The fact that you have more than one computer affected strongly indicates that there is a common problem (i.e., do you have the same A/V or firewall on those PCs? You'll obviously be using the same network).

What does the firewall log say for the period when Boinc is trying to do an upload? What does the A/V log say? What appears in the Boinc messages log? (Note that one of the upload servers went down on Friday evening).

The following workunit shows no significant problems apart from being aborted by user:
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7176043

The messages you are referring to are unrelated to this network problem (they're warning messages, not error messages). I think 'CPDN Monitor - Quit request from BOINC...' just means that you're shutting down the PC.

This one shows no problems apart from being aborted via transfer abort:
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7079461

This one shows some issues, but not fatal ones (until it was aborted by user, that is):
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=6265777

Note that upload issues won't appear on these pages, because zip file uploads are a separate process.

Your first computer:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_host_detail.php?hostid=798233

Your computer list:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/hosts_user.php?userid=93294&show_all=1&sort=rpc_time

Are you sure 707181 is yours? This is linked to a different account.
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_host_detail.php?hostid=707181

There is a section in the CPDN READMEs for network issues:
http://www.climateprediction.net/board/viewtopic.php?t=5896

---

And also see this post which talks about problems with Norton (affecting project communication and also file exclusive-locking issues).

http://www.climateprediction.net/board/viewtopic.php?p=38884#38884

---

The computers are 816152 & 798233. The other computer seems to have worked itself out. I have these work units stacked up like cordwood...5 on one and 3 on the other...maybe I am just getting them done too fast. I started another project to divert some of the wasted time. But they are constantly trying to upload with no success.
I shut off the firewall completely and retried the uploads...no success.

This is the latest from my wbemess, I have no clue what it means but maybe it will make sense to you.

(Sun Jan 13 10:02:44 2008.68054140) : Failed to log an event: 5DE
(Sun Jan 13 10:02:44 2008.68054140) : Dropping event destined for event consumer NTEventLogEventConsumer="SCM Event Log Consumer" in namespace //./root/subscription
(Sun Jan 13 10:02:44 2008.68054140) : Failed to deliver an event to event consumer NTEventLogEventConsumer="SCM Event Log Consumer" with error code 0x80041001. Dropping event.

This is what the message reads in Boinc Manager:

1/13/2008 1:23:43 PM|climateprediction.net|Started upload of hadam3h_n_098s5_009c_009c_0_0_4.zip
1/13/2008 1:27:33 PM||Project communication failed: attempting access to reference site
1/13/2008 1:27:33 PM|climateprediction.net|Temporarily failed upload of hadam3h_n_098s5_009c_009c_0_0_4.zip: http error
1/13/2008 1:27:33 PM|climateprediction.net|Backing off 1 hr 10 min 2 sec on upload of hadam3h_n_098s5_009c_009c_0_0_4.zip
1/13/2008 1:27:35 PM||Access to reference site succeeded - project servers may be temporarily down.

And also see this post which talks about problems with Norton (affecting project communication and also file exclusive-locking issues).

http://www.climateprediction.net/board/viewtopic.php?p=38884#38884

Hi Sharlee

I can only see current 2 workunits/tasks on computer #798233 and 1 on computer #816152.

Nobody should be without firewall protection even for a moment. I hope that before you disabled your Norton firewall you checked that the Windows firewall was enabled.

The 'HTTP error' makes me suspect that this maybe isn't your firewall problem, Sharlee, but a problem that's been occurring intermittently with the CPDN upload servers in Oxford for over a month. Have a look at the CPDN News thread here, specifically the post on 24 Dec:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=5447#32153

If this is the case, the only solution is, as that post says, to keep trying at intervals.

Can you tell us how exactly how many zip files appear in the Transfers window of each computer and if possible how long they have been sitting there?

I'm now not so sure that your problem is that the BOINC servers are misbehaving. A CPDN trickle of mine has just been refused by the server, but the message was

14/01/2008 15:40:39|climateprediction.net|Scheduler request failed: HTTP internal server error

whereas your message just said 'http error'. So perhaps it is a connection problem from your end after all.

Sharlee
You said in your first post:

They stay stuck in my computer in upload.

Just to be sure that I've got this right; does that mean that there are some zip files in the Transfers tab? What is their approx size? (It may be a file size limit in some node between you and Oxford Uni.)

What is the size of files for the other projects that ARE working? (You may have to set Network option to off for a while so that you can see them.)
(AOL has (or did have) a file size limit.)

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

Also, some info on messages may help:
Do you keep getting retry count downs in the Projects tab?
What is the Status in Tasks?
What is the failure message in Messages?
(Always http error? As Mo said, in 5.10.28 it SHOULD be HTTP internal server error if it's a server error.)

The files aren't that big that they shouldn't go through. I tried one of the suggestions to put the http://bbc.cpdn.org in the advanced section, options, HTTP Proxy, all the files went out instantly(5 of them size between 1.5 an 4.8)and they are now listed as complete. On another computer I had 3 files and this worked there as well. I had one problem doing this and that was with a rosetta file that was complete as well, it became corrupted. If anyone chooses this option be sure to update your other projects first. Insert the proxy address listed above and when the files go...change the setting back to the way it was before.

Thanks everyone for all the help!

Sharlee
You said in your first post:

They stay stuck in my computer in upload.

Just to be sure that I've got this right; does that mean that there are some zip files in the Transfers tab? What is their approx size? (It may be a file size limit in some node between you and Oxford Uni.)

What is the size of files for the other projects that ARE working? (You may have to set Network option to off for a while so that you can see them.)
(AOL has (or did have) a file size limit.)

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

Also, some info on messages may help:
Do you keep getting retry count downs in the Projects tab?
What is the Status in Tasks?
What is the failure message in Messages?
(Always http error? As Mo said, in 5.10.28 it SHOULD be HTTP internal server error if it's a server error.)