I have 5 active (boinc) pc's in my farm that were running fine until last weekends outage. When everything came back up, 1 started getting this message. Absolutly nothing changed on my end. The machine is Win2K and using 4.35 and runs 24/7. It has an always on connection in a large network
(yes, the connection is working). The 2 computers next to it are crunching classic and have not missed a beat before or since. There is also a Win 2K machine running 4.19 in the next room that is fine. I have been through everything and nothing will work. As of right now, it is sitting idle with all of the WU's "ready to report" and no way to update or anything. I am ready to reset the project but I have noticed several reports of this recently.

---

have you tried exiting the client and restarting it again?
Have you tried exiting the client, rebooting the computer and starting it again?

Check your Windows Event Viewer, see if there's a Master Browser or other similar fight going on. That may kill your network connection. Even though it may not kill your internet connection, it may interfere.

---

Here's another thought. Not an answer but It may narrow it down a bit. First lets assume his broadband connection to the internet is working. I'm assuming this because:

He's able to post to this thread
he's able to a couple computers reach Classic Seti
He's able to reach Boinc with his computer using Boinc 4.19

So, the connection from his router/ISP to the internet is working.

So, What's different? It appears it's the 4.35 Boinc Manager and it's Client. I don't know if the Software Firewall if present is different from the others. We know the 4.19 client is NOT a GUI, and the 4.35 client IS a GUI with CLI, or maybe just CLI.

I know that with 4.19 only the CLI requires my firewall to "allow access".
I know with 4.20-4.38 "Access" is required by BOTH the GUI and the CLI. Perhaps your 4.35 manager/client haven't both (2) gotten the "Allow Access" from the software firewall?

tony

PS I haven't noticed others listing this problem that are using 4.35 specifically. The other thread is using 4.36. Nor have I seen a rash of reports with this problem since 4.35 was released. So it seems more likely the problem is on your end. That's what I'd check.

Who says he posted this from the PC that was affected?

---

<blockquote>Who says he posted this from the PC that was affected? </blockquote>


Hmmm, could be.

Maybe the power is off to the stack and the terminal server is getting that message?

Maybe the aliens have broken the traces on his PC so he couldn't find them?

1. I have restarted and rebooted
2. I have uninstalled and tried 4.32, 4.25, 4.37 and 4.19 with no change.
3. I had completed 12 WU's with 4.35 with no problems before the outage
4. In addition I have 2 boxes running classic and 1 running 4.19 on this LAN
5. While these posts came from home, I have had full internet access from this computer and have been able to check my stats and network status from this pc.
6. one thing I have not done but will try is to see if I can dl a classic wu tomorrow.
7. this is computer # 606675
8. All of this started with last weekends outage
9. could this be a bad or corrupt line in one of the xml files?

---

Bump, this still isn't resolved. I really need some new ideas for this one.
FYI, I have been able to run classic now for 4 days with no issues.
I tried to clean install 4.19 and 4.42 today with the same result of "Master File fetch failed" and then it detaches.
I have checked everything that I can think of with no progress.
Win 2K, SP4
# 606675

---

have you checked the software firewall "internet applications" and made sure you have both "Boincmgr.exe, AND Boinc.exe" listed to allow connection?

tony

I'd say to Check this thread for more options, especially on your network and firewall!

Check if the Win2k machine uses a software firewall. See what specific ports are open on all the other machines. See what ports are open on the router you use to have all these PCs connect to the internet simultaniously.

---

No one has added a firewall to this machine. I have 2 others plus now this one running classic and one running 4.19 running through the local switch.
I was able to update the project and get my finished work reported but this machine could not/did not get the reply from Berkeley It could/ did upload work but could not get replies.
Berklely even assigned fresh work on the last update but my local BOINC never knew it. This box can transmit but not receive on BOINC.
It is fully functional with classic and the internet.

---

Is port 1043 open? Boinc version 4.19 and earlier uses 31416 and the later versions use 1043. I'm not sure which port Classic uses. This port issue is a possibility. I'm not sure how to manually check them as I use a firewall program.

tony

through the local switch

(emphasis mine)

How is this network connected to the internet? Work from the switch where the PCs are attached to, back to where the data flows into the wall-outlet, please.

---

I will have to check the ports tomorrow

---

Are they PCs at work?

---

Yes, these 4 are at work. I do have permission to use them. They run 24/7 just in case somebody needs to to use them because it would take to long to boot them and the network. 99% of the time they are sitting there doing absolutly nothing.

---

Yes, these 4 are at work. I do have permission to use them. They run 24/7 just in case somebody needs to to use them because it would take to long to boot them and the network. 99% of the time they are sitting there doing absolutly nothing.

SEVERAL people have reported that the problem went away when they put info i nthe proxy server area of Boinc. Try it, see if it works, if it doesn't then so be it, BUT if it does....you are home free!

---

Yes, these 4 are at work. I do have permission to use them. They run 24/7 just in case somebody needs to to use them because it would take to long to boot them and the network. 99% of the time they are sitting there doing absolutly nothing.

SEVERAL people have reported that the problem went away when they put info i nthe proxy server area of Boinc. Try it, see if it works, if it doesn't then so be it, BUT if it does....you are home free!

Anything in particular I should put there?

---

Yes, these 4 are at work. I do have permission to use them. They run 24/7 just in case somebody needs to to use them because it would take to long to boot them and the network. 99% of the time they are sitting there doing absolutly nothing.

SEVERAL people have reported that the problem went away when they put info i nthe proxy server area of Boinc. Try it, see if it works, if it doesn't then so be it, BUT if it does....you are home free!

Anything in particular I should put there?

Any of these should work:
195.229.240.86 PORT : 80
82.101.132.51 PORT : 8080
62.193.231.243 PORT : 8080
195.229.240.86 PORT : 80
193.188.96.19 PORT : 80
193.188.96.138 PORT : 80
217.17.233.181 PORT : 80
66.160.69.101 PORT : 80
202.56.253.183 PORT : 8080
The connection will NOT be as fast, but should work. These came up during the recnt problems with servers overseas not connecting.

---

Hi. I'm not sure if you have fixed this now but I had a similar experience recently when installing a optimised client. My win 2003 box (used only for seti) used my Linux box for DNS. I had taken down the DNS a few days earlier and restarted it. It appears that when 2003 restarted it had lost its address for DNS queries and just kept reporting it could not connect to seti@home. Took some time to fathom what was happening but as soon as I corrected the problem all went well. So if you are still having problems it might be worth checking out all the network components. Make sure the domain name is being resolved using ping. Make sure you have a route to the dta servers - use trace route. Hope this helps as it sound similar...all I did was re-boot and nothing else had changed....or so I thought. Not sure at all why win 2003 lost the address for dns though.

---