Windows machine hangs while uploading

Message boards : Number crunching : Windows machine hangs while uploading
Message board moderation

To post messages, you must log in.

AuthorMessage
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1963644 - Posted: 6 Nov 2018, 16:28:43 UTC

So i've been having an intermittent problem with one of my windows machines.

Basically what seems to be happening is intermittently a completed WU begins uploading, it hangs, and then all of the other WU after do the same thing. it never changes to "Ready to report", and stays in the state of "Uploading" or "preparing to upload" whatever the current verbiage is. it stays like this until it runs out of work, as it wont download new work saying that there are too many tasks currently uploading.

there is no problem with the system's internet connection. everything else works just fine.
it will sit like this for DAYS if i dont catch it, so a temporary server issue shouldnt be the cause.
I only see this on ONE system. my other 2 windows systems, one with very similar hardware, do not have this issue at all.
restarting BOINC doesn't work.
the only fix i've found is rebooting the entire PC. then it picks right back up like nothing happened
there are no errors in the BOINC log, other than saying it failed to upload and it was backing off, but it does this continually until a reboot. fixes it.

i thought the issues was maybe BOINC was corrupted or something, i had this issue happening on BOINC 7.12.1, and hoped a platform update would remedy it, but after updating, the problem persists.

basic system specs:
Windows 7 x64 w/ all latest updates
BOINC 7.14.2
Supermicro X9DAE w 2x E5-2697v2
2x GTX 1080ti

this one : https://setiathome.berkeley.edu/show_host_detail.php?hostid=8433872

Does anyone know a more concrete reason why this is happening? I'm hoping this is a known problem and someone can point me in the right direction, I'd rather not resort to a full system wipe. It seems like "something" is getting hung up somewhere, but in a really transparent way preventing upload to SETI even though there are no issues with the network connection.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1963644 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1963650 - Posted: 6 Nov 2018, 16:49:13 UTC - in response to Message 1963644.  

Twice I have had 1 tasks get stuck at Ready to Report, but it kept ticking right along with the 1 tasks sitting there. Nothing cures that but editing the client_state to get it going again. I believe both times this was a network interruption.

Does restarting BOINC work, or just a full reboot? I'm curious if a Network Stop/Start would work.
ID: 1963650 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1963657 - Posted: 6 Nov 2018, 21:04:48 UTC - in response to Message 1963650.  

Twice I have had 1 tasks get stuck at Ready to Report, but it kept ticking right along with the 1 tasks sitting there. Nothing cures that but editing the client_state to get it going again. I believe both times this was a network interruption.

Does restarting BOINC work, or just a full reboot? I'm curious if a Network Stop/Start would work.


restarting BOINC does not work. only a full system reboot seems to fix it (so far since i havent narrowed down the exact cause yet).

when you say network stop/start, is this a function in BOINC? or do you mean something like disabling and enabling the network adapter via Windows settings?
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1963657 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1963662 - Posted: 6 Nov 2018, 21:19:13 UTC - in response to Message 1963657.  

Easiest it to first try the stopping of Network comms via the Manager. Wait out the 5 minute interval and re-enable net comms. I've done that several times to unstick a "stuck" upload. It works. If that doesn't it is just as quick to reboot the whole PC rather than fumbling with disabling a network adapter and then enabling again.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1963662 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1963673 - Posted: 6 Nov 2018, 22:16:09 UTC - in response to Message 1963662.  

Thanks Keith, I'll give that a try next time it pops up. Seems to be happening every week or so randomly.

The system in question is "remote" to me, so it's usually several hours before i notice it, then drive over to fix it.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1963673 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1964050 - Posted: 9 Nov 2018, 3:57:56 UTC

Well today I noticed 2 systems hung, both windows machines, both on very similar hardware (very similar motherboards). When i checked them, they both were exhibiting the same issue, hung upload tasks that were filling the queue and blocking the download of new work.

So i disabled/enabled the Network adapter in Windows settings (not BOINC), and doing that, BOINC picked up right away and started going again, i didnt even have to restart BOINC. but it's weird, neither system was disconnected from the internet when i looked at them, verified with bringing the web browser up, but BOINC still couldnt see anything. It's like if there is a temporary network interruption, BOINC gets blinded so to speak and can no longer see the internet, even though it's there.

both motherboards have dual LAN ports, so i moved the Ethernet cable from one port to the other port, to see if that will change anything. probably not, but worth trying to rule it out. Next will be wiping out the network drivers and re-installing the ones from the manufacturer. last resort will be an OS re-install or just swapping over to Linux.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1964050 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1964053 - Posted: 9 Nov 2018, 4:15:00 UTC

Next time you catch them doing that, set http_debug and http_xfer_debug in the Event Log Options. Look for a message that says Seti DNS address STALE - zapped. That is probably the problem. When you enable the debug, it forces the client to refresh the connection. Look for DNS issues with the host. Try adding another DNS server besides the default one for your ISP.

Also try and define the Seti server IP addresses in the hosts file instead of letting your default DNS servers try and figure it out. I have those saved into a document in my Seti docs folder in case I need to edit the hosts file. You can just leave the entries in the file and put a comment# in front of them to ignore them normally. And uncomment them when you are having DNS issues.

208.68.240.118 setiboincdata.ssl.berkeley.edu # upload server Oct 2016
208.68.240.119 boinc2.ssl.berkeley.edu # Georgem download server Oct 2016
208.68.240.126 setiboinc.ssl.berkeley.edu # scheduler Oct 2016
208.68.240.127 vader.ssl.berkeley.edu # Vader download server Oct 2016
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1964053 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1964065 - Posted: 9 Nov 2018, 5:15:32 UTC

A simple network suspend/resume in BOINC Manager should reset the connection.

I leave the DNS entries Keith mentioned in my hosts file. I can remove them if it ever causes problems, but haven't had to do that.
ID: 1964065 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1964066 - Posted: 9 Nov 2018, 5:24:25 UTC - in response to Message 1964065.  

Honestly I couldn’t find it’s location in BOINC. That’s why I just tried it in windows.

But it’s still not a great solution. Since the systems are remote. I have to drive over to their location to push the buttons. Would rather it just not do this in the first place lol.

I’ll add the lines in the host file if this keeps happening.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1964066 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1964080 - Posted: 9 Nov 2018, 6:40:23 UTC - in response to Message 1964066.  

?? You couldn't find the location for network suspend in the Manager?

From the Activity menu selection. >> Network activity suspended

??For the location of the hosts file?

/etc/hosts

Needs to be edited as root.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1964080 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1964153 - Posted: 9 Nov 2018, 16:23:54 UTC - in response to Message 1964080.  

Thanks Keith about the network setting location within BOINC. I only briefly looked and didnt want to spend too much time on it so I went to the Windows side network settings first. I'll give it a shot next time it pops up.

about the host file, this is a Windows system, so no /etc/ directory in root hahaha.

but i'm familiar with the host file on Windows systems. the correct location is C:\Windows\System32\drivers\etc\hosts

Windows uses backslash (\), Unix/Linux uses forward slash (/) for directory structure.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1964153 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1964155 - Posted: 9 Nov 2018, 16:30:38 UTC - in response to Message 1964153.  

You should look at installing TeamViewr on your remote Windows computers for desktop access to them - it works great.
It is also available for Linux, but I haven't tried it yet. Anyone use it??

I use VNC locally for my Linux boxes. But TeamViewer is much more router friendly for remote systems.
ID: 1964155 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1964160 - Posted: 9 Nov 2018, 16:46:38 UTC

I wont touch TV with a 10ft pole.

I'm in the process of building my own VPN server with OpenVPN and running it on the NAS at this location. I do the same thing at my house and it works great once it's all setup. I just need the spare time to finish the configuration.

OpenVPN runs in a jail on FreeNAS. Once I'm in, I can just use VNC.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1964160 · Report as offensive
Profile Siran d'Vel'nahr
Volunteer tester
Avatar

Send message
Joined: 23 May 99
Posts: 7379
Credit: 44,181,323
RAC: 238
United States
Message 1964161 - Posted: 9 Nov 2018, 16:49:09 UTC - in response to Message 1964155.  

You should look at installing TeamViewr on your remote Windows computers for desktop access to them - it works great.
It is also available for Linux, but I haven't tried it yet. Anyone use it??

I use VNC locally for my Linux boxes. But TeamViewer is much more router friendly for remote systems.

Hi Brent,

Wasn't there a thread here about TV where they were pulling something like a Micro$oft in trying to get money from users of the free version by switching it to Enterprise or some such thing? Users had to jump through hoops with TV in order to continue to use the free version and it was time consuming.

Have a great day! :)

Siran
CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath
ID: 1964161 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1964221 - Posted: 9 Nov 2018, 22:22:14 UTC - in response to Message 1964161.  

Yes, some people have had problems with TV saying - Are you sure this is for personal use, pay up then ...
ID: 1964221 · Report as offensive
Dr Who Fan
Volunteer tester
Avatar

Send message
Joined: 8 Jan 01
Posts: 3219
Credit: 715,342
RAC: 4
United States
Message 1964305 - Posted: 10 Nov 2018, 5:24:08 UTC - in response to Message 1964161.  

Yep, TeamViewer has been pulling a M$ trick of trying extort $$$ out of a turnip.

The most recent reports from TV users on TV forum site indicate it is getting (slightly) better... TV users that have hit by the pay to use"feature" and got a reset received a messages from TV support indicating TV can only be used to connect to/from one residential IP address to another residential IP unless you want to pay for a commercial licence.

In other words, TV says you NOT to use it on your laptop at you favorite coffee shop using their FREE WIFI, at school, etc to connect to your own home PC without a commercial licence version for BOTH PC's.
ID: 1964305 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1964569 - Posted: 11 Nov 2018, 6:08:37 UTC

I seen something when poking around ... a command line option.

--file_xfer_giveup_period N
Specify giveup period for file transfers

Maybe of some use to test???
ID: 1964569 · Report as offensive

Message boards : Number crunching : Windows machine hangs while uploading


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.