Can\'t report to server on one machine

Message boards : Number crunching : Can\'t report to server on one machine

To post messages, you must log in.

AuthorMessage
Plasmon_attack

Send message
Joined: 2 May 10
Posts: 13
Credit: 15,451,384
RAC: 0
Message 69649 - Posted: 16 Feb 2011, 3:13:35 UTC

Hi Everyone, recently one of my computers (Win7 64-bit, dual quad core xeons, 6 GB ram, hyperthreaded, boinc 6.10.58) suddenly started having problems communicating with the server. It\'s still crunching work but says, \"Scheduler request failed: Couldn\'t connect to server.\" Other computers on the same network are working fine.

I\'ve uninstalled and reinstalled, made sure to delete the Boinc directory, and yet when I reinstall it\'s like nothing happened because the same work units are ready to report, the project is still attached, etc.

The error is, \"Boinc couldn\'t do internet communication, and no default connection is selected. Please connect to the internet, or select a default connection using advanced/options/connection.\"

I don\'t need a proxy or anything for my network. Any idea what to put in? Is there a better way to reinstall? I have like 300 completed workunits piled up and would like to report them.

Thanks!
Tony
ID: 69649 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
transient
Avatar

Send message
Joined: 30 Sep 06
Posts: 376
Credit: 8,798,867
RAC: 0
Message 69652 - Posted: 16 Feb 2011, 8:37:44 UTC

Could it be the firewall on that particular computer is blocking the connection?
ID: 69652 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
mikey
Avatar

Send message
Joined: 5 Jan 06
Posts: 1449
Credit: 5,776,642
RAC: 0
Message 69653 - Posted: 16 Feb 2011, 11:28:37 UTC - in response to Message 69649.  

Hi Everyone, recently one of my computers (Win7 64-bit, dual quad core xeons, 6 GB ram, hyperthreaded, boinc 6.10.58) suddenly started having problems communicating with the server. It\'s still crunching work but says, \"Scheduler request failed: Couldn\'t connect to server.\" Other computers on the same network are working fine.

I\'ve uninstalled and reinstalled, made sure to delete the Boinc directory, and yet when I reinstall it\'s like nothing happened because the same work units are ready to report, the project is still attached, etc.

The error is, \"Boinc couldn\'t do internet communication, and no default connection is selected. Please connect to the internet, or select a default connection using advanced/options/connection.\"

I don\'t need a proxy or anything for my network. Any idea what to put in? Is there a better way to reinstall? I have like 300 completed workunits piled up and would like to report them.

Thanks!
Tony


Does the pc otherwise connect to the net just fine? And you are seeing the exact same units because that is a \'feature\' that was turned on recently here at Rosie, it has been at some other projects for awhile now. If you are using a wireless connection try running a temporary wire to the machine and see if that fixes it.
ID: 69653 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mod.Sense
Volunteer moderator
Project administrator

Send message
Joined: 22 Aug 06
Posts: 3520
Credit: 0
RAC: 0
Message 69655 - Posted: 16 Feb 2011, 14:04:37 UTC

Have you rebooted the PC?

Have you tried to PING boinc.bakerlab.org from that machine? (perhaps the machine is having trouble reaching a DNS server to resolve the domain name)
Rosetta Moderator: Mod.Sense
ID: 69655 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Plasmon_attack

Send message
Joined: 2 May 10
Posts: 13
Credit: 15,451,384
RAC: 0
Message 69660 - Posted: 16 Feb 2011, 19:25:22 UTC - in response to Message 69655.  

Yes, I tried rebooting first, and several times, and between installations of different versions of the BOINC client.

It appears that our network admins may have disabled ping as I can\'t ping anything (even Google) from any computers here (note one is a laptop and ping works fine at home). Note, the network is working as there are seven other computers on the same network that are able to reach the server without an issue.

There was a power outage that crashed the machine (yes it\'s on a backup but the outage was too long) and this computer hasn\'t been able to connect.

I guess, is there a better way to uninstall Boinc? Clearly the uninstall doesn\'t do it totally because, once reinstalled, it knows it\'s connected to Rosetta AND still knows what work units are completed and not completed. Is there a way to wipe it off completely and start over?

Thanks
ID: 69660 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
TPCBF

Send message
Joined: 29 Nov 10
Posts: 108
Credit: 3,695,073
RAC: 2,793
Message 69662 - Posted: 16 Feb 2011, 20:46:30 UTC - in response to Message 69660.  

Yes, I tried rebooting first, and several times, and between installations of different versions of the BOINC client.

It appears that our network admins may have disabled ping as I can\'t ping anything (even Google) from any computers here (note one is a laptop and ping works fine at home). Note, the network is working as there are seven other computers on the same network that are able to reach the server without an issue.

There was a power outage that crashed the machine (yes it\'s on a backup but the outage was too long) and this computer hasn\'t been able to connect.

I guess, is there a better way to uninstall Boinc? Clearly the uninstall doesn\'t do it totally because, once reinstalled, it knows it\'s connected to Rosetta AND still knows what work units are completed and not completed. Is there a way to wipe it off completely and start over?

Thanks
I had the same problem with one of my laptops. All the sudden it wouldn\'t connect back to the server, getting the same error messages you mentioned. Nothing changed on the laptop itself, nothing changed on the LAN/firewall, which would effect all of my roughly one dozen PCs crunching. I had to uninstall BOINC (running only R@H) and manually deleted both the C:\\Program files\\BOINC folder as well as the C:\\Documents and Settings\\All Users\\Application Data\\BOINC folder. Can it be you deleted only the first one but not the second one (where the \"Application Data\" folder is a \"hidden\" folder)? Once I did that, I was able to re-install (actually downloaded the binary installer from World Community Grid) and re-attach to the R@H project. Lost the three finished WU in the process though...

This was on Windows XPSP3 though, on Windows 7, the second folder should of course be in \"C:\\Users\\...\"

Ralf
ID: 69662 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Murasaki
Avatar

Send message
Joined: 20 Apr 06
Posts: 303
Credit: 481,795
RAC: 73
Message 69663 - Posted: 16 Feb 2011, 20:57:45 UTC

A few things to try:


  • Check that the computer has a network connection - are you able to use an internet browser/send emails/view files on other machines on the network?
  • In BOINC \"Advanced View\" check whether you have \"Network activity always available\" selected in the Activity menu.
  • Attach the machine to another BOINC project and see if you can communicate with them.


ID: 69663 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
TPCBF

Send message
Joined: 29 Nov 10
Posts: 108
Credit: 3,695,073
RAC: 2,793
Message 69664 - Posted: 17 Feb 2011, 0:34:01 UTC - in response to Message 69663.  

As I mentioned, I had the very same issue and in my case
*Check that the computer has a network connection - are you able to use an internet browser/send emails/view files on other machines on the network?
Network and Internet connection was just fine, no problems what so ever. Only BOINC had connectivity issues, for whatever reason...
*In BOINC \"Advanced View\" check whether you have \"Network activity always available\" selected in the Activity menu.
Didn\'t make a difference. All other clients have the default of \"...based on preferences\" and work just fine
*Attach the machine to another BOINC project and see if you can communicate with them.
Catch22, you can\'t do that when BOINC claims that it doesn\'t have Internet connectivity.

It\'s not a R@H specific issue, rather a BOINC related issue. I tried at first to re-install the client downloaded from the World Community Grid and it wouldn\'t work either.

As I mentioned before, I had to uninstall the program as well as delete the application data folder before a re-install would work just fine as if nothing ever happened.

This is definitely a snag in the BOINC software, as it happened to my on a host that was doing at that time nothing but just crunching for R@H and out of nowhere experienced those issues...

Ralf
ID: 69664 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Plasmon_attack

Send message
Joined: 2 May 10
Posts: 13
Credit: 15,451,384
RAC: 0
Message 69665 - Posted: 17 Feb 2011, 8:54:57 UTC - in response to Message 69664.  

Umm, yes, I did check that the computer has internet access and that BOINC is allowed to do network access before asking for help.

@Wolf...your situation and suggestions are a lot closer to what I think the issue is, namely, uninstall doesn\'t fully uninstall and leaves \'ghosts\' behind that keep a fresh install from functioning. I\'ve tried what you said and am a little further, but not quite there. I was able to locate program data in a the \'program data\' folder. I also found a leftover directory in program files (x86). ProgramData has the project specific data, work units, a lot of .xml files, etc., and I wiped it out. A fresh install now prompts to setup a new project.

However, it still can\'t communicate with the servers. It says they\'re temporarily unavailable (but I tried several projects) and then it gives the same network error.

For Win7 are there any other hidden places the program could be leaving traces that remind if of this broken state? I\'ve looked in the users directories (and I have hidden fiels showing) but I\'m not finding anything that looks boinc related.

Sorry about your 3 workunits. I\'ve been trying to fix this for a while as all 16 cores were crunching and lost ~400 workunits :(
ID: 69665 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dcdc

Send message
Joined: 3 Nov 05
Posts: 1615
Credit: 59,053,137
RAC: 40,700
Message 69666 - Posted: 17 Feb 2011, 9:00:04 UTC - in response to Message 69665.  

There are a few registry entries, but I don\'t think they contain anything that would affect network access - AFAIK they just contain the locations of the boinc folders, but MS would be able to clarify that.

Could the firewall have been set to block BOINC?

Umm, yes, I did check that the computer has internet access and that BOINC is allowed to do network access before asking for help.

@Wolf...your situation and suggestions are a lot closer to what I think the issue is, namely, uninstall doesn\'t fully uninstall and leaves \'ghosts\' behind that keep a fresh install from functioning. I\'ve tried what you said and am a little further, but not quite there. I was able to locate program data in a the \'program data\' folder. I also found a leftover directory in program files (x86). ProgramData has the project specific data, work units, a lot of .xml files, etc., and I wiped it out. A fresh install now prompts to setup a new project.

However, it still can\'t communicate with the servers. It says they\'re temporarily unavailable (but I tried several projects) and then it gives the same network error.

For Win7 are there any other hidden places the program could be leaving traces that remind if of this broken state? I\'ve looked in the users directories (and I have hidden fiels showing) but I\'m not finding anything that looks boinc related.

Sorry about your 3 workunits. I\'ve been trying to fix this for a while as all 16 cores were crunching and lost ~400 workunits :(


ID: 69666 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Plasmon_attack

Send message
Joined: 2 May 10
Posts: 13
Credit: 15,451,384
RAC: 0
Message 69667 - Posted: 17 Feb 2011, 9:10:39 UTC - in response to Message 69666.  

Ok, got it, I worked backwards through all versions of the BOINC client and found that 6.6.38 works. I\'m back up, though I think the server is mad at me for losing the ~400 units that were completed so it only gave me one to start with (for 16 cores!). I guess eventually it\'ll recover.
ID: 69667 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Murasaki
Avatar

Send message
Joined: 20 Apr 06
Posts: 303
Credit: 481,795
RAC: 73
Message 69669 - Posted: 17 Feb 2011, 23:27:15 UTC - in response to Message 69665.  
Last modified: 17 Feb 2011, 23:28:10 UTC

Umm, yes, I did check that the computer has internet access and that BOINC is allowed to do network access before asking for help.


When someone asks for technical assistance there are two choices. One, assume it is something complex and start with the most complicated solutions. Two, decide to rule out the obvious possibilities first.

Going with the second option does result in some people thinking you are an idiot for suggesting the obvious, but it is surprising how often the simple things are overlooked.
ID: 69669 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dcdc

Send message
Joined: 3 Nov 05
Posts: 1615
Credit: 59,053,137
RAC: 40,700
Message 69673 - Posted: 19 Feb 2011, 11:24:59 UTC - in response to Message 69667.  

Ok, got it, I worked backwards through all versions of the BOINC client and found that 6.6.38 works. I\'m back up, though I think the server is mad at me for losing the ~400 units that were completed so it only gave me one to start with (for 16 cores!). I guess eventually it\'ll recover.

good to hear it\'s up and running again ;)
ID: 69673 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : Can\'t report to server on one machine



©2019 University of Washington
http://www.bakerlab.org