Message boards : Number crunching : Upload of WU yes......Upload of Ready to Report no
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Hathor
Avatar

Send message
Joined: 17 Sep 04
Posts: 27
Credit: 298,995
RAC: 0
Message 2299 - Posted: 25 Sep 2004, 20:40:01 UTC


My program just started this problem a few of days ago.

The finished WU's upload when done in a flash.

The Ready to Report WU's will not upload with the message: Temporarily failed upload of (insert WU here)

I upgraded to 4.09, no help, then re-installed 4.05 still no help.

Any Ideas? I have 73 RtR with two added every hour.





<a> [/url]
ID: 2299 · Report as offensive     Reply Quote
Holmis

Send message
Joined: 17 Sep 04
Posts: 10
Credit: 5,620
RAC: 0
Message 2302 - Posted: 25 Sep 2004, 21:04:11 UTC

Redy to report means that BOINC is waiting to report to the scheduler that the work is done and uploaded. Try to manually update the project under the projects tab.
ID: 2302 · Report as offensive     Reply Quote
Profile Stephen Balch

Send message
Joined: 22 Sep 04
Posts: 42
Credit: 2,956
RAC: 0
Message 2305 - Posted: 25 Sep 2004, 21:20:39 UTC

When I find WU's for any project in the "Read to report" status. I always do a manual update. I don't wait for it to get around to sending the "Rtr" status automatically. It makes me happy because I have really completed the process for a result, it makes others, who may be waiting for my result as the third of a set of results, happy because we should soon get credit for that WU, and it may allow LHC to do any further processing on the results more quickly.

Stephen
ID: 2305 · Report as offensive     Reply Quote
Hathor
Avatar

Send message
Joined: 17 Sep 04
Posts: 27
Credit: 298,995
RAC: 0
Message 2308 - Posted: 25 Sep 2004, 21:44:29 UTC


I have done that over and over.

Force Update from the Projects Tab and Force Retry now on the Transfer Tab.

No Joy, and a bunch of them are _3
ID: 2308 · Report as offensive     Reply Quote
Profile Stephen Balch

Send message
Joined: 22 Sep 04
Posts: 42
Credit: 2,956
RAC: 0
Message 2318 - Posted: 26 Sep 2004, 2:00:48 UTC

It's my understanding the "Ready to report" status results are reported automatically when the application determines it needs to request more work and contacts the scheduler. It seems to for me, if I haven't been manually updating because I'm in class.

BTW, I'm running BOINC 4.09 on an AMD Athlon XP 3100+ (1.7GHz, not over clocked) with Win 98SE, with SETI 4.03 and LHC 4.45. Since you are running Windows XP Home, I would recommend you change back to BOINC 4.09, or the current "latest public version," rather than continue to run 4.05.

Your message text, "Temporarily failed upload of (insert WU here)," makes it sound more like you are trying to a upload result than report a result status, when it fails. Are you sure you are actually getting result files uploaded to the servers?

Back when SETI was down a lot, I would get a lot of sequences like: "Started upload of xxxxx", followed by "Temporarily failed upload of xxxxx", followed by a "Backing off m minutes and s seconds on transfer of file xxxxx" when trying to contact the server to upload results, not report results. It would try another contact m minutes and s seconds later. It failed more often than not. Eventually, it would make contact, but at the time contacting SETI was very hit or miss. Now that SETI is, mostly, back up, I don't get those messages much any more.

When the application tries to contact the scheduler to send a status, but no schedulers responds, you will get an "Scheduler RPC to xxxxx failed" error, an "No schedulers responded" error, followed by a "Deferring communication" message. It then retrys after the defferal period expires.

The schedulers at SETI@home and LHC@home seem to be up, at least I've been able to upload results and report status to both of them today. My last upload to SETI was at 2004-09-25 15:47:48, and my last report of an "Rtr" was at 2004-09-25 16:28:50. I last uploaded a result to LHC at 2004-09-25 16:35:02, and reported a "Rtr" result at 2004-09-25 18:35:15. Both uploads went without a hitch, and both reports reported "succeeded".

As I type this message (2004-0925 20:26), BOINC has requested and received WU's from both SETI and LHC, and downloaded sixtrack_4.46_windows_intelx86.exe from LHC. I still have about 20 WU's in queue that show sixtrack 4.45 as their processor, and now 24 WU's showing 4.46. (Dates and times are U.S. Central Daylight Time (CDT), UTC - 5 hours.)

Are you sure you have the correct URLs for the projects? That they haven't been corrupted somehow?

Cheers,

Stephen



ID: 2318 · Report as offensive     Reply Quote
Hathor
Avatar

Send message
Joined: 17 Sep 04
Posts: 27
Credit: 298,995
RAC: 0
Message 2322 - Posted: 26 Sep 2004, 2:59:32 UTC


I have two computers running LHC both for the same amount of time. One is uploading WU's as they finish and uploading the ready to report automatically.

The other, uploads the finished WU's as they are finished, however every time the ready to reports are sent (automatically or using retry) I get the error.

According to the Transfer Tab, their is a connection at xx.xxKB, yet every time the upload fails.

Updating from the Projects Tab fails, Retry from the Transfer Tab fails, and no new WU's are being downloaded or the account info updated.

Nothing has changed on the computer that is failing, same connection, firewall, etc. This problem occured out of the blue.

Very frustrating.

ID: 2322 · Report as offensive     Reply Quote
Profile Stephen Balch

Send message
Joined: 22 Sep 04
Posts: 42
Credit: 2,956
RAC: 0
Message 2385 - Posted: 26 Sep 2004, 18:18:19 UTC - in response to Message 2322.  

Kevin,

One obvious difference I see is that one of your computers is an Intel P4 (2 CPUs), while the other is an AMD Athlon(tm) XP 2600+ (1 CPU). Which processor is failing on reporting? Could you be more specific?
>
> I have two computers running LHC both for the same amount of time. One is
> uploading WU's as they finish and uploading the ready to report
> automatically.
>
> The other, uploads the finished WU's as they are finished, however every time
> the ready to reports are sent (automatically or using retry) I get the error.
>
The average transfer rates also seem to be significantly different. The AMD Athlon(tm) XP 2600+: uploads at 6.06 KB/sec, and downloads at 268.72 KB/sec. The Intel Pentium 4, with 2 CPUs: uploads at 0.65 KB/sec, and downloads at 147.11 KB/sec, yet you say they use the same connection. Is the P4 the one that is failing reporting? Do you use a shared dial up/a cable modem/LAN connection? Have you checked your firewall log for blocking of one the transfers?

Not to mention the measured processing rates, the AMD Athlon(tm) XP 2600+: measured floating point speed 1976.83 million ops/sec, measured integer speed 4508.54 million ops/sec, and the Intel Pentium 4, with 2 CPUs: measured floating point speed 1735.81 million ops/sec per CPU, measured integer speed 2374.25 million ops/sec per CPU.

The point being that the two computers are not identical, and I'm trying to determine what the difference between the two is. We already know the calculation results will probably be different because of the processor types, perhaps significantly different. Are any of the WinXP components CPU-type specific? I don't know much about XP, not yet running it myself.
>
> Very frustrating.
>
Computers can be that way. (GRIN)

Cheers,

Stephen

ID: 2385 · Report as offensive     Reply Quote
Hathor
Avatar

Send message
Joined: 17 Sep 04
Posts: 27
Credit: 298,995
RAC: 0
Message 2410 - Posted: 26 Sep 2004, 23:27:42 UTC
Last modified: 26 Sep 2004, 23:28:10 UTC

@ Stephen Balch (et al)

The AMD is not having any problems at all. It is the P4 that cannot connect.

I have four computers total connected to an aDSL modem through a Netgear RP 614. All computers can access the internet and can upload and download without a problem.

I just removed all Norton Security from the P4 thinking that might be the problem even-tho Boinc_gui.exe was allowed automatic access to the internet. That is the same setting I still have on the AMD box.

Now with Norton removed, forcing an update to the project returns no schedulers responded.

The WU's on the Transfer Tab now connect (as before) except instead of uploading at xx.xxKB with no progress indicated, they connect upload approx 47% then fail (everyone, every time).

The connection to the internet is good, I can ping, connect to the LHC website, all things about the connection are good, except the client can not connect to the server to update or upload successfully.

I have made no changes to the P4 box since building it and installing BOINC then attaching to LHC. As you might have noticed, the P4 Last contacted the server 24 Sep 2004 8:52:59 UTC and was created about 11 minutes after the AMD.

I will continue to tweak the box, however I have no clue what to do about the client that would somehow force it to co-operate. Maybe their is something in the ini files that is messed up? Sending the wrong information (handshake if you will)


To make matters worse, I just built another AMD box with a clean install of Win XP and it will not connect to LHC to download my account, and my other AMD box with S@H BOINC installed is having the same problems. It worked flawless for months before encountering this problem. I have scanned all for virus spy ware and such, no joy.

Anybody have a magic bullet for this?





ID: 2410 · Report as offensive     Reply Quote
Profile Krunchin-Keith [USA]
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 2 Sep 04
Posts: 209
Credit: 1,482,496
RAC: 0
Message 2411 - Posted: 27 Sep 2004, 0:19:11 UTC
Last modified: 28 Sep 2004, 18:10:37 UTC

ID: 2411 · Report as offensive     Reply Quote
Profile meckano
Avatar

Send message
Joined: 17 Sep 04
Posts: 150
Credit: 20,315
RAC: 0
Message 2413 - Posted: 27 Sep 2004, 0:32:45 UTC - in response to Message 2299.  
Last modified: 27 Sep 2004, 0:34:08 UTC

Maybe check,
Your account/Boinc General settings: There is a place that says to connect to network every 0.1 days; I figured it had something to do with that.
- If they are still there after a day, and you have a permanent internet connection, maybe the server has been busy when you connected. But this part never happened to me.

Well, it happened with seti, but they said they had problems.
ID: 2413 · Report as offensive     Reply Quote
Hathor
Avatar

Send message
Joined: 17 Sep 04
Posts: 27
Credit: 298,995
RAC: 0
Message 2424 - Posted: 27 Sep 2004, 3:32:39 UTC


Thanks to all that replied.

It seems that my router turned on me like a rabid dog. Once I connected directly to my aDSL modem and went through the hoop to release and make a new connection to my ISP for each box everything worked out.

It still puzzles me that I could connect to the internet and all, but the BOINC client (and windows activation it turns out) could not find the connection!!

Now all I have to do is make the router friendly again, so I do not have to log on and see that special deal for Home Networking at only $99.00 for hardware and $9.99 additional per month for the Honor. Shoot, at almost $60.00 a month now they have to be crazy.

I will post back when I get things sorted out.

Again, thank you all for the replies.




<a> [/url]
ID: 2424 · Report as offensive     Reply Quote
Profile Stephen Balch

Send message
Joined: 22 Sep 04
Posts: 42
Credit: 2,956
RAC: 0
Message 2430 - Posted: 27 Sep 2004, 6:51:00 UTC

Kevin,

I'm glad to read you figured it out and got (or are in the process of getting) it worked out.

Cheers,

Stephen
ID: 2430 · Report as offensive     Reply Quote
Hathor
Avatar

Send message
Joined: 17 Sep 04
Posts: 27
Credit: 298,995
RAC: 0
Message 2520 - Posted: 28 Sep 2004, 1:21:11 UTC
Last modified: 28 Sep 2004, 1:22:36 UTC

Well, I have a new problem now.

The hooking up each box to the aDSL modem worked fine then I started to have a problem with the P4 again.

2004-09-27 18:08:09 - May run out of work in 5.00 days; requesting more

2004-09-27 18:08:10 - Can't resolve hostname www.lhcathome.cern.ch (valid name, no data record of requested type)

LHC@home - 2004-09-27 18:08:10 - Couldn't read master page for LHC@home: error -113

LHC@home - 2004-09-27 18:08:10 - Master file fetch failed

LHC@home - 2004-09-27 18:08:10 - Deferring communication with project for 1 days, 1 hours, 17 minutes, and 31 seconds

and if I try to do a manual update on the Transfer Tab, it just ignores me and keeps on counting down to the next retry.

Now what? The other boxes hooked up to LHC work like a charm.

ID: 2520 · Report as offensive     Reply Quote
Darrell

Send message
Joined: 18 Aug 04
Posts: 8
Credit: 126,861
RAC: 0
Message 2522 - Posted: 28 Sep 2004, 1:31:29 UTC

LHC@home - 2004-09-27 18:08:10 - Couldn't read master page for LHC@home: error -113

This is the rather cryptic line that boinc prints out when it can not reach the internet. Since this is a new box connection and all the other boxes worked, suggest you check this ones connections again.
ID: 2522 · Report as offensive     Reply Quote
Hathor
Avatar

Send message
Joined: 17 Sep 04
Posts: 27
Credit: 298,995
RAC: 0
Message 2523 - Posted: 28 Sep 2004, 2:03:26 UTC

@ Darrell

That is what is so annoying. The connection works fine to access the internet (I am using the P4 now), I can do a windows update, check E-mail, and use other programs that require access to the net.

It is the BOINC client (v 4.09) that is giving me fits.

It has firewall permission for outbound and inbound communications and it worked fine last night.

I have made no changes to the connection or the OS, nor have I installed any new software.

This is becoming a real head banger!

What if I copied all of the files from the BOINC folder Projects and uninstalled then reinstalled the client. After which I moved the Projects folder back to the "clean" install. Anybody think this would do the trick?

Thank you all again for your time and suggestions.
ID: 2523 · Report as offensive     Reply Quote
Profile Stephen Balch

Send message
Joined: 22 Sep 04
Posts: 42
Credit: 2,956
RAC: 0
Message 2530 - Posted: 28 Sep 2004, 4:16:16 UTC - in response to Message 2523.  

Kevin,

That's one of the reasons we're here reading these threads. (GRIN)

I don't think it's BOINC 4.09 or the LHC client. Darrell is right about the LHC message, as far as it goes, and obviously you can access the internet from that P4, the real question is, "Why can't the LHC client access the LHC web site?"

The more interesting line, to me, is:

2004-09-27 18:08:10 - Can't resolve hostname www.lhcathome.cern.ch (valid name, no data record of requested type

This sounds more like a DNS problem to me. Is your ISP having DNS problems? It seems to recognize the LHC URL, but it can't find the IP address record for LHC. It is trying to convert the URL you gave it to the IP address it really needs.

Do you have to point to the proper DNS IP at your ISP, or do you pick up your networking parameters from your ISP?

Cheers,

Stephen
ID: 2530 · Report as offensive     Reply Quote
Holmis

Send message
Joined: 17 Sep 04
Posts: 10
Credit: 5,620
RAC: 0
Message 2532 - Posted: 28 Sep 2004, 5:28:17 UTC - in response to Message 2520.  


> 2004-09-27 18:08:10 - Can't resolve hostname www.lhcathome.cern.ch (valid
> name, no data record of requested type)
It seems to me that it's the wrong adress, I got this adress when I signed up.

http://lhcathome.cern.ch

What adress did you attach to???
If other than my above, try this instead...

ID: 2532 · Report as offensive     Reply Quote
Hathor
Avatar

Send message
Joined: 17 Sep 04
Posts: 27
Credit: 298,995
RAC: 0
Message 2544 - Posted: 28 Sep 2004, 8:47:04 UTC


I am back up and running but at a cost.

I saved the projects folder then uninstalled 4.09. After reinstalling and attaching to the project I was able to connect and download another group of WU's which have been crunching and uploading automatically as it should be. That is the good news.

The bad news is when I moved the saved files (ready to reports and unfinished WU's) into the new projects folder, they were not recognized by the client.

At least the master page and master file fetch problem is solved and it was not my connection that was causing the mayhem.

I am glad that the finished WU's were uploaded, sorry for anyone that was waiting on the ready to reports.

I believe their had to be some kind of corruption in one of the client files that caused the problems. Too bad their was no way of editing the offender and making things right.

Thank you all once again. I hope this is the end of it and I will not have to post yet again.

ID: 2544 · Report as offensive     Reply Quote
eberndl

Send message
Joined: 27 Sep 04
Posts: 5
Credit: 1,358,998
RAC: 0
Message 2625 - Posted: 28 Sep 2004, 21:43:16 UTC

You might want to check if LHC now has 2 computers listed on your system. If you DO have 2 comps, merge them and you might once again have access to your older WUs... just a thought.
ID: 2625 · Report as offensive     Reply Quote
Holmis

Send message
Joined: 17 Sep 04
Posts: 10
Credit: 5,620
RAC: 0
Message 2638 - Posted: 28 Sep 2004, 22:34:07 UTC - in response to Message 2625.  

> You might want to check if LHC now has 2 computers listed on your system. If
> you DO have 2 comps, merge them and you might once again have access to your
> older WUs... just a thought.
>

If the units aren't in the *.xml files the client will not recognize them.
ID: 2638 · Report as offensive     Reply Quote
1 · 2 · Next

Message boards : Number crunching : Upload of WU yes......Upload of Ready to Report no


©2024 CERN