Message boards : Number crunching : Connection problems
Previous · 1 · 2 · 3 · 4 · 5 · Next
Author | Message |
---|---|
Moderator9 Volunteer moderator Send message Joined: 22 Jan 06 Posts: 1014 Credit: 0 RAC: 0 |
Mod9, he IS working again, I emailed him the Rosetta app, and he used that. He could download the app from other projects, but no matter when he tried or how (either manually downloading from site, or automatically during attachment), he couldn't get the app file downloaded from Rosetta. He can download all the other stuff from Rosetta, but not the LARGE file. He's was getting connection dropped or cut off. There are a few reports of NO APP file downloading. I am leaning toward the problem being on Rosettas server/file handler end. Probably in how it negotiates a transaction, some time limit or something. Tony, Thanks. I will bring your thoughts on this to the system admins attention. He might be able to tweak something. Moderator9 ROSETTA@home FAQ Moderator Contact |
JPLiz Send message Joined: 2 Mar 06 Posts: 41 Credit: 215,433 RAC: 0 |
Or, you say it's working OK on one of your puters, you can copy it from the working one, and paste it to the other.no no no I just tried predictor to see if it work and it does but no rosetta yet. Yes thanks I am up and running now. Thanks to Tony. As for the speed,i don't think this is the issue. I have a DSL an able to D/L large files from other programs AKA peer to peer. For now i am ok but i would like to be inform if you guys fine the prob. Thanks again for all the help, Your's truly JP. |
JPLiz Send message Joined: 2 Mar 06 Posts: 41 Credit: 215,433 RAC: 0 |
Mod9, he IS working again, I emailed him the Rosetta app, and he used that. He could download the app from other projects, but no matter when he tried or how (either manually downloading from site, or automatically during attachment), he couldn't get the app file downloaded from Rosetta. He can download all the other stuff from Rosetta, but not the LARGE file. He's was getting connection dropped or cut off. There are a few reports of NO APP file downloading. I am leaning toward the problem being on Rosettas server/file handler end. Probably in how it negotiates a transaction, some time limit or something. Hey how are you hihi They have updated again and i still have the same prob. Could you send me the update via email plz? I also have the same pron whit seti so if you have that one too i would lukre to have if my email is able to receive 10 meg in one message. my email is jpaube at hotmail dot com. Thanks and i'm still working on the prob here. I'm sepose to get a new moden in a few days an see if this is going to help if not i'm going to change ISP. Hope u had fun at the race later JP. |
Astro Send message Joined: 2 Oct 05 Posts: 987 Credit: 500,253 RAC: 0 |
JP....You've got mail MOD9 anyone figure why these people can't complete downloads yet? Have they or are they looking into it? tony |
Moderator9 Volunteer moderator Send message Joined: 22 Jan 06 Posts: 1014 Credit: 0 RAC: 0 |
JP....You've got mail The System administrator looked into it and he can't find any problem on the server side. From what he can tell the system is sending the files, but the client system does not "ack" the completed download. From what I can see, it looks like the client is dropping the conection for some reason. Moderator9 ROSETTA@home FAQ Moderator Contact |
JPLiz Send message Joined: 2 Mar 06 Posts: 41 Credit: 215,433 RAC: 0 |
JP....You've got mail Hey MOD9 Mi ISP is sending a tech here Friday to check the connection and the modem to see if this is the cause. It's pozzeling everybody even senior tech at my ISP so i'll let you know if they fine something. -:) Have a nice day JP. |
JPLiz Send message Joined: 2 Mar 06 Posts: 41 Credit: 215,433 RAC: 0 |
JP....You've got mail Ok my isp was here and all is ok on this side he came whit a laptop an we where able to get norton software ouver 10 mgs exe file but he was unable to get the rosetta exe. Everything here is ok but i srill cant get the exe from rosetta. For now it's ok because i had it email to me but if there is another update i will eather have to stop the priject or get it emailed to me again. BTW he was not on my network but he was connected in the modem. Later JP. |
Moderator9 Volunteer moderator Send message Joined: 22 Jan 06 Posts: 1014 Credit: 0 RAC: 0 |
JP....You've got mail It is a shame he could not test from inside your network. From the Rosetta side there does not seem to be a problem. Clearly if you dial in directly there is no problem. So I suspect something in you network setup. I presume you are using a configurable switch for this purpose. Have you looked at the configuration parameters for your switch? Moderator9 ROSETTA@home FAQ Moderator Contact |
Astro Send message Joined: 2 Oct 05 Posts: 987 Credit: 500,253 RAC: 0 |
mod9, what he's saying is that the tech couldn't get the rosetta file either, using his laptop and JP's modem. The only thing left is the modem, the ISP, the backbone providers and Rosetta. |
Moderator9 Volunteer moderator Send message Joined: 22 Jan 06 Posts: 1014 Credit: 0 RAC: 0 |
mod9, what he's saying is that the tech couldn't get the rosetta file either, using his laptop and JP's modem. The only thing left is the modem, the ISP, the backbone providers and Rosetta. Ok, I misread the post. It still does not look like the Rosetta server is the issue. Teh Sys Admin looked for a problem and found nothing in that end. Moderator9 ROSETTA@home FAQ Moderator Contact |
BennyRop Send message Joined: 17 Dec 05 Posts: 555 Credit: 140,800 RAC: 0 |
If it was a problem with the Rosetta server, then there should be more of us affected. And if it was a problem with Jean-Paul's modem, then he shouldn't be able to download a 10 meg file from Symantec. I find it strange that the "tech" didn't run a tracert to the Rosetta server to identify where the possible problems are. When tracking down problems playing various online games, they'll usually give plenty of directions on finding out who is at fault. |
Feet1st Send message Joined: 30 Dec 05 Posts: 1755 Credit: 4,690,520 RAC: 0 |
I find it hard for the tech to say it's not the fault of the ISP, when he can't get the download to complete either. Seems pretty clear to me, that since there are not more people reporting such a problem that it IS the ISP. I mean there's only two PCs (I think?) that we've heard having such a problem, one is JP's and the other is JP's tech... and BOTH were using the same ISP! Coincidence?? JP, do you have any ISPs there that will give you a 30 day free trial? Perhaps you could try another ISP, or use a friends account or something just to see if you can prove there's something specific to that ISP's servers and routers. Add this signature to your EMail: Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might! https://boinc.bakerlab.org/rosetta/ |
Astro Send message Joined: 2 Oct 05 Posts: 987 Credit: 500,253 RAC: 0 |
A user by the name of Finch also reported similar problems earlier in this thread here. It that thread he states: I have nothing to contribute to a solution, but wanted to say you aren't the only one. There do not seem to be a significant number of people with this problem because I have never found much on this in forums, newsgroups, etc. I have this problem with lhc and with rosetta. I can run einstein@home and world computing grid(which also uses rosetta) without problems. I can download work units with no trouble. When I try to download the application file for these two projects it craps out at some point. Sometimes at 90+% through the transfer, sometimes as little as 20%. I have tried to deal with this off and on for a year or more using different versions of boinc or newer versions of the application. I have never been able to solve it. Running on xandros and dial-up. Since this doesn't seem to be a common enough problem I think we are out of luck. I am glad to see that you too have found some boinc projects that you can take part in. I've kept JP going by emailing him the file, but the other small minority might just give up instead of working to use Rosetta. I wonder if Management couldn't contact JP via his listed email address and get JP's IP addy from him, so they could check the Apache logs to see where it's being dropped. JP might even use Ethereal and try to trace it himself, or atleast be able to send management the log. JP might also try to borrow someone elses modem to rule his out as the culprit. Just because JP's a member of an extremely small minority doesn't mean he shouldn't be allowed to participate, or that his concerns aren't valid. tony |
Moderator9 Volunteer moderator Send message Joined: 22 Jan 06 Posts: 1014 Credit: 0 RAC: 0 |
...JP might also try to borrow someone elses modem to rule his out as the culprit. I don't believe anyone associated with the project has suggested that just because the problem is limited to two reports that these people will be abandoned. I have certainly not suggested that. AS I said the sys admin has looked that this and cannot find any problem on the server side. While not by any means a suggestion that these two people do not deserve help, it is significant that there are not more reports. The clear implication is that there is something not right on the remote end of the connection. IT could very well be that the e-mails are compressed in transmission and fall below some maximum file limit. The fact that software from other projects that are using Rosetta software can be downloaded, is not significant. None of those project is using the current (and larger) Rosetta application. I will again draw attention to this thread to the sys admin, but I suspect that the modem or the ISP are the source of the issue. It could be that the modem overheats with a high sustained load and shuts off, it could be that the ISP has a limit the technician is not aware of. It could be a lot of things, but we will keep at it until we find the cause. Moderator9 ROSETTA@home FAQ Moderator Contact |
BennyRop Send message Joined: 17 Dec 05 Posts: 555 Credit: 140,800 RAC: 0 |
When I had the problems I was talking about with an online game - I was told to try a different ISP.. and to run Tracert (pingplotter would give even better information, since it can be run for hours and show network results over time.) to see where problems were happening. ACS worked fine with the game. GCI got all the way to Seattle, and there was a high rate of packet loss. I then had to call GCI's tech support, have them run a tracert from their network to the online game server - and they saw the same problem. Then they dealt with the corrupted router in Seattle that was causing the problem. Assuming (often dangerous..) that the download site for the new rosetta updates is https://boinc.bakerlab.org (feel free to post the actual ip# and url for the client update site - and the site used by the automatic updates..) - I get this from running "tracert https://boinc.bakerlab.org >c:rosetta.txt" and then opening Rosetta.txt with a text editor. Tracing route to https://boinc.bakerlab.org [66.150.161.133] over a maximum of 30 hops: 1 52 ms 50 ms 51 ms 66-230-xxx-xxx-dsl-rb1.kdk.acsalaska.net [66.230.xxx.xxx] You might want to remove the first hop.. or mask out the ip# with xxx's like I've done. 2 52 ms 52 ms 51 ms fe1-0-104-cr1.kdk.acsalaska.net [209.193.35.105] 3 52 ms 51 ms 52 ms 209.193.6.17 4 692 ms 703 ms 683 ms 209.193.6.25 5 694 ms 680 ms 704 ms 209.193.6.26 6 699 ms 710 ms 706 ms ge6-2-gsr1.nwc.acsalaska.net [209.112.165.140] 7 723 ms 735 ms 740 ms s3-2-gsr1.sea.acsalaska.net [209.193.35.249] 8 715 ms 734 ms 742 ms sl-gw16-sea-2-4.sprintlink.net [144.228.93.21] 9 717 ms 707 ms 729 ms 144.232.6.163 10 733 ms 724 ms 747 ms sl-gw12-sea-1-0.sprintlink.net [144.232.6.122] 11 716 ms 733 ms 735 ms sl-internap-127-0.sprintlink.net [144.228.93.66] 12 730 ms 717 ms 741 ms border5.ge5-1-bbnet1.sef.pnap.net [63.251.160.10] 13 720 ms 735 ms 741 ms redirectf.dnsix.com [66.150.161.133] Trace complete. What this tells us: There's 13 steps between me and Rosetta.. that in my case, the first three hops are on island, and the rest have gone over a satellite jump which will hopefully be eliminated by Oct of this year.. :) If you have major problems with lost packets, you'll see astericks "*" instead of a time like my gawd-awful 741ms. If the problem is being caused by packet loss, then Tracert or PingPlotter will show you where the problem lies. |
finch Send message Joined: 23 Nov 05 Posts: 8 Credit: 4,548 RAC: 0 |
Everyone keeps talking about this as if it has to be either something on the server end OR something on our end. People seem to forget that this is a very rare problem on our end as well. If it was an obvious problem on my end it would be causing me much more extensive nuisance. I don't think it is a heat issue. I have run this modem for 7 years, in 80+ F mid summer indoor temps downloading hundreds of megabytes of linux distro for 10-12 hours at a time. I'm skeptical as to it being some conscious limit my isp is engaging in. They have always been very hands off, just providing a connection and staying out of my way to let me use it(quite intensively and never a complaint from them. I in turn don't have any significant complaints about them). I have never seen evidence of aggressive controlling behavior in their network management, so I doubt they have been overly zealous in configuring something that is causing trouble. I think it is more likely that there is some subtle interaction of factors at BOTH the server end and my isp that only rarely crops up. If it was simply one OR the other I don't think the issue would be as rare at the server end, if it was a server problem, or as rare at my end, if it was an isp or system problem. |
BennyRop Send message Joined: 17 Dec 05 Posts: 555 Credit: 140,800 RAC: 0 |
I pointed out a way to see if it was network "in between" - on one of the many routers/servers between the end user, and the Rosetta Server. One misconfigured router near Seattle spelled doom for my connection to a gaming server at least 5 years ago. I later noticed that many of the Cisco routers used on the saming server's isp were running old versions of IOS that were susceptible to certain attacks - i.e. it was possible to get the affected routers to create enough traffic on the ISP to knock it off the internet. (The gaming server's ISP went bankrupt..) |
finch Send message Joined: 23 Nov 05 Posts: 8 Credit: 4,548 RAC: 0 |
I pointed out a way to see if it was network "in between" - on one of the many routers/servers between the end user, and the Rosetta Server.) traceroute looks normal to me every hop is like these, with 140-200ms. 13 uwbr-ads-01-te3-1.cac.washington.edu (209.124.176.23) 189.844 ms 189.994 ms 190.262 ms 14 acar-ads-02-vlan3998.cac.washington.edu (140.142.155.25) 189.711 ms 189.245 ms 210.032 ms 15 boinc.bakerlab.org (140.142.20.103) 199.897 ms !C 199.926 ms !C 199.968 ms !C mtr run for 30 secs or so a few times during evening hours shows 2% loss on one hop and none on the other 14 hops, average 200 ms or less on all hops, worst 500ms. Nothing looks troubling to me. Granted this wasn't run near the time of my problem downloading, but I have had this problem over many months of trying at various times. |
JPLiz Send message Joined: 2 Mar 06 Posts: 41 Credit: 215,433 RAC: 0 |
I find it hard for the tech to say it's not the fault of the ISP, when he can't get the download to complete either. Seems pretty clear to me, that since there are not more people reporting such a problem that it IS the ISP. I mean there's only two PCs (I think?) that we've heard having such a problem, one is JP's and the other is JP's tech... and BOTH were using the same ISP! Coincidence?? Yes i'll try that thanks |
JPLiz Send message Joined: 2 Mar 06 Posts: 41 Credit: 215,433 RAC: 0 |
This is my trace to the BOINC network. Does this look ok to you? To me it dose so where is the prob?...JP might also try to borrow someone elses modem to rule his out as the culprit. 1 0 0 0 66.98.244.1 gphou-66-98-244-1.ev1.net 2 0 1 0 66.98.241.16 gphou-66-98-241-16.ev1.net 3 4 1 1 66.98.240.14 gphou-66-98-240-14.ev1.net 4 1 1 1 129.250.11.141 ge-1-3-0.r02.hstntx01.us.bb.gin.ntt.net 5 52 52 52 129.250.2.1 p16-1-1-0.r21.plalca01.us.bb.gin.ntt.net 6 53 53 53 129.250.5.49 p64-0-0-0.r21.mlpsca01.us.bb.gin.ntt.net 7 70 70 70 129.250.4.22 p64-0-0-0.r20.sttlwa01.us.bb.gin.ntt.net 8 70 70 70 129.250.2.205 xe-3-1.r00.sttlwa01.us.bb.gin.ntt.net 9 70 70 70 129.250.10.193 ge-0.uw.sttlwa01.us.bb.gin.ntt.net 10 68 66 67 209.124.176.23 uwbr-ads-01-te3-1.cac.washington.edu 11 67 67 67 140.142.155.25 acar-ads-02-vlan3998.cac.washington.edu 12 67 66 66 140.142.20.103 boinc.bakerlab.org |
Message boards :
Number crunching :
Connection problems
©2024 University of Washington
https://www.bakerlab.org