A follow on to my previous problem (~2-3 days ago) where it uploaded stuff that it never started processing.

It was continually requesting 0 seconds of work from CPDN. ProteinPredictor@Home was doing the same so clearly it was a Boinc (4.45) problem. I dropped both projects and readded them (after resetting the project didn\'t fix the problem).

Anyway, after adding CPDN back in it downloaded work units (at least it looked like it did), however they never showed up as work to do.

According to the log:
The app suite downloaded the following WU:
6/11/2005 11:05:03 PM|climateprediction.net|Finished download of 05zk_000032774.zip

Later that day it downloaded another (why???):
6/11/2005 11:33:16 PM|climateprediction.net|Finished download of 05i1_100032136.zip

Neither ever showed up in Boinc and if I check the Results section it gives \"Download Error,\" yet both (according to Boinc) downloaded.

Results links: <a>05zk...</a>, <a>05i1...</a>.

If you look at those error messages they aren\'t with the data but instead with the HADSM download (not finished), yet your end is giving me a \"Daily Quota Reached\" message so I can\'t get them.

05zk file xfer errors (stripping XML stuff):
app_version download error: couldn\'t get input files:
hadsm3data_4.12_windows_intelx86.zip
hadsm3um_4.12_windows_intelx86.zip

05i1 file xfer errors:
app_version download error: couldn\'t get input files:
hadsm3um_4.12_windows_intelx86.zip

Second one had managed to get the data part I guess.

Anyway, I\'m confused? Is Boinc 4.45 not able to successfully hold a download until the app that processes it is downloaded?

(I\'m on a 56k modem)

--RC

Are there any extra messages in your log files (eg stderr.txt or stderr.old or stdout.txt/old)?

Hopefully now it has got hadsm3data_4.12_windows_intelx86.zip it won't try again. If you continue to have trouble getting hadsm3um_4.12_windows_intelx86.zip maybe a ftp download would work but I am not sure about it needing to be signed. It isn't on your system already, is it?

The following suggests that other messages could be possible:
2/28/2005 7:38:31 PM|climateprediction.net|Scheduler RPC to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi succeeded
2/28/2005 7:38:32 PM|climateprediction.net|Started download of hadsm3data_4.04_windows_intelx86.zip
2/28/2005 7:38:32 PM|climateprediction.net|Started download of hadsm3se_4.04_windows_intelx86.zip
2/28/2005 7:38:38 PM|climateprediction.net|Finished download of hadsm3se_4.04_windows_intelx86.zip
2/28/2005 7:38:38 PM|climateprediction.net|Throughput 152294 bytes/sec
2/28/2005 7:38:38 PM|climateprediction.net|Started download of hadsm3um_4.04_windows_intelx86.zip
2/28/2005 7:38:38 PM|climateprediction.net|signature verification error for hadsm3se_4.04_windows_intelx86.zip
2/28/2005 7:38:38 PM|climateprediction.net|Checksum or signature error for hadsm3se_4.04_windows_intelx86.zip
2/28/2005 7:38:38 PM|climateprediction.net|Unrecoverable error for result 2idq_200138320_1 (app_version download error: couldn't get input files: hadsm3se_4.04_windows_intelx86.zip -120 signature verification error)
2/28/2005 7:38:39 PM|climateprediction.net|Deferring communication with project for 59 seconds
2/28/2005 7:38:51 PM|climateprediction.net|Finished download of hadsm3um_4.04_windows_intelx86.zip
2/28/2005 7:38:51 PM|climateprediction.net|Throughput 187540 bytes/sec

STDERRAE (thats for Boinc Manager...Boinc normal is STDERR) contains interesting info about where the problem occured. From what I can tell is that BOINC Manager has real problems when the client isn't online. Basically the following is the most useful part:

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

2005-06-11 22:20:40 [---] Can't resolve hostname [climateapps2.oucs.ox.ac.uk] host name not found
2005-06-11 22:20:40 [climateprediction.net] Couldn't start download of 05zk_000032774.zip
2005-06-11 22:20:40 [climateprediction.net] URL http://climateapps2.oucs.ox.ac.uk/cpdnboinc/download/05zk_000032774.zip: error -113
2005-06-11 22:20:40 [climateprediction.net] Unrecoverable error for result 05zk_000032774_0 (app_version download error: couldn't get input files:
&lt;file_xfer_error&gt;
&lt;file_name&gt;hadsm3data_4.12_windows_intelx86.zip&lt;/file_name&gt;
&lt;error_code&gt;-200&lt;/error_code&gt;
&lt;error_message&gt;&lt;/error_message&gt;
&lt;/file_xfer_error&gt;
&lt;file_xfer_error&gt;
&lt;file_name&gt;hadsm3um_4.12_windows_intelx86.zip&lt;/file_name&gt;
&lt;error_code&gt;-200&lt;/error_code&gt;
&lt;error_message&gt;&lt;/error_message&gt;
&lt;/file_xfer_error&gt;
)
2005-06-11 22:20:41 [climateprediction.net] Deferring communication with project for 58 seconds

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

There is no reason why it should return an error of this sort on a "Can't resolve host name" - I think the Boinc devs need to be advised that not everyone has 24/7 internet...and even those that do sometimes get knocked offline. I don't want to give Boinc control over my modem...I have one phone line and I don't want it dialing if I'm expecting a call!

I can give you the full dump for CPDN from the current messages (haven't shut down Boinc since then) or from the STDERRAE file. I can Grep them so you don't have to hunt through tons of useless stuff for other projects.

See
http://www.climateprediction.net/board/viewtopic.php?t=2887&amp;start=105

Same error messages -113. This is normally associated with firewall or other communication issues, but we know you got the data.zip file :?

Les Bayliss thinks there is another problem at Oxford. (That wouldn't suprise me either.)


I wonder if instead of creating misallocated WU, we now get failures. That could be considered an improvement, but that probably isn't much consolation to you Random Chaos.

Hope they can get things running smoothly again on Monday.

My Boo-Boo Crandles.
I obviously wasn't fully awake. At 5.30am I went to upload an overnight trickle and found that I hadn't disabled net access, so there were lots of red error messages. BUT I wasn't logged in, so the errors were BOINC not getting to the net.

I just 'woke up', (pun intended), to the problem and tried again, with success.

Sorry all.
(Now, where else did I post?)

Les

I agree it would make sense to try again rather than trash the WU without trying again.

However before rushing to that conclusion, I wonder what a -200 error is? I don't think I have seen one of those before. Anyone seen a list of these errorcode numbers?

&gt; I agree it would make sense to try again rather than trash the WU without
&gt; trying again.
&gt;
&gt; However before rushing to that conclusion, I wonder what a -200 error is? I
&gt; don't think I have seen one of those before. Anyone seen a list of these
&gt; errorcode numbers?
&gt;

Well, it was definately a network issue becuase my modem dropped the connection and I didn't dial back in for about an hour. During that time it gave the -200 error - perhaps that is something like "network not found"? Whatever it is Boinc shouldn't drop a WU becuase it can't find the server...it should check whether the network is even there before doing anything beyond giving a notice.

Frankly there should be different colors between a "Warnings" (things like server connection issues and no work available) and a "Fatal Error" (WU failed), as well as an option to turn off "unable to connect" type messages so that us modem users don't get spammed...but that is for the Boinc dev team that seems to ignore complaints of its users about their app's functionality problems.

I have to say...I've been running Seti@Home since 2000 and didn't have any problems with it until they switched to Boinc. Ever since then I've had regular problems with the app, from taking too much CPU time to the sort of thing this thread is about.

--RC

Crandles
There was a list of error codes on the community forum. I think the poster said something like "here is an old list of error codes I had saved".
I noticed that the messages about UOTW have disappeared about Feb-March, which is probably due to reloading the files from a backup after the hacker attack, so the error codes message may be among them. I've wondered about them myself, so I guess that I should spend some time looking for it.

Les

Found it within a minute of reading your post:

<a href="http://boinc.berkeley.edu/error_numbers.h">here</a>
from
http://www.climateprediction.net/board/viewtopic.php?t=2416&amp;highlight=list+error+codes

-200 = wrong size

Thanks Les

Is there a safety issue if the file is the wrong size? A bit like signature verification error, it may be dangerous to leave such unsigned files on your PC, so for safety sake, it failed the whole thing? (I still think delete it and try again would be better.)

Wrong size sort of makes sense. The file was partially (or not at all) downloaded when the modem cut off. Clearly 0 != correct size!

When the modem cuts off it "should" resume downloading next time I connect. Looking back at the log it did do this:

<b>6/11/2005 11:39:32 PM|climateprediction.net|Started download of hadsm3um_4.12_windows_intelx86.zip</b>
6/11/2005 11:39:32 PM|climateprediction.net|Unrecoverable error for result 05i1_100032136_0
(app_version download error: couldn't get input files:&lt;file_xfer_error&gt; &lt;file_name&gt;hadsm3um_4.12_windows_intelx86.zip&lt;/file_name&gt; &lt;error_code&gt;-200&lt;/error_code&gt; &lt;error_message&gt;&lt;/error_message&gt;&lt;
6/11/2005 11:39:33 PM|climateprediction.net|Deferring communication with project for 58 seconds
6/11/2005 11:40:32 PM|climateprediction.net|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi
6/11/2005 11:40:32 PM|climateprediction.net|Requesting 8640 seconds of work, returning 1 results
6/11/2005 11:40:47 PM|climateprediction.net|Scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi succeeded
6/11/2005 11:40:47 PM|climateprediction.net|Message from server: No work sent
6/11/2005 11:40:47 PM|climateprediction.net|Message from server: (reached daily quota of 2 results)
<b>6/11/2005 11:44:55 PM|climateprediction.net|Finished download of hadsm3um_4.12_windows_intelx86.zip</b>
6/11/2005 11:44:55 PM|climateprediction.net|Throughput 7245 bytes/sec


Interestingly it gives the error on "05i1_100032136_0" WHILE hadsm3um_4.12_windows_intelx86.zip was downloading...um...and the error is that hadsm3um is the wrong size?

Of course it is the wrong size...it is still downloading!

Crazy Boinc devs :)

------
PS: Someone really needs to allow either span color or font color to work on the BBS...hard to highlight lines without that.

And just an updated note:

It finally downloaded a WU and is ready to run...after 3 days of nothing and screwy errors the couple days before that. Lets hope Boinc doesn't have any more odd application logic.

Glad you found it, crandles. By the time I'd finished looking through the other boards, I'd forgotten about it.

Maybe my memory needs defragmenting. :)

Hope you get there this time RC.

Thanks for all your help, Les and crandles.

&gt; Glad you found it, crandles. By the time I'd finished looking through the
&gt; other boards, I'd forgotten about it.
&gt;
&gt; Maybe my memory needs defragmenting. :)
&gt;
&gt; Hope you get there this time RC.