Hughesnet Community

Gen5 ftp broken since 15 December 2018

cancel
Showing results for 
Search instead for 
Did you mean: 
mweasner
Sophomore

Gen5 ftp broken since 15 December 2018

Since 15 December 2018 ftp access to my web server has stopped working reliably day and night. Sometimes I can get a reliable fast connection, other times a connection but very slow, and many times no connection at all.  I tried a friend's ftp server that uses a different host in a different part of the country and still have the ftp problem.  I searched the Forum and see that over the years many users have had ftp problems with HughesNet.  http access to the server works fine.  I have rebooted the HN modem and one of the computers used to access ftp but that didn't help. I see this problem on multiple Macs and multiple iOS devices so it would appear that the common source of the problem is HughesNet.  Why?  What's the fix to get reliable HN ftp again?

54 REPLIES 54

Hello,

 

Jumping in for Jay while he's out the next couple of days. I won't dance around the fact that I am not happy to hear what occurred on your call to both our support lines. Agreed with Mark on most points, but the time setting could be a recommendation to make sure that there are no conflicts with certificates and that any handshakes don't fail because of time discrepancies. Most likely it is all in regard to security protocols being touchy.  I'll check in with you tomorrow to get updates on the situation.

 

Thanks

Amanda

Update after 24 hours with the different DNS and bandwidth settings.  Ethernet connected computers have always connected to the ftp server on multiple tries over the past 24 hours.  Wi-Fi connected devices had success once but have failed multiple times over the past 24 hours.  I have not yet rebooted the modem; waiting to see if the Ethernet connected computers (which are primary for ftp access) lose their ftp access.

Have also noticed that loading of web pages in the web browser is much snappier than it has been for weeks.

Not ready to declare success yet but looking more hopeful, thanks to the settings changes that the Home Technical Support rep made yesterday.

That is great news! As for the wireless devices, do you have a way to test these devices for packet loss over the air? Do the devices exhibit the same behavior on both 2.4GHz and 5GHz?

Sorry for the delay in updating on the ftp issue.  After changing the DNS setting on the modem on Thursday, 21 March, it appeared that the ftp connection issue might have been resolved, at least for the Ethernet connected computers.  I made several ftp connections on Thursday and Friday from multiple computers with different OS versions and they connected without any problems.  Saturday morning, 23 March, I connected OK from one computer, however Saturday afternoon the connection from the same computer to the same ftp server failed.  Over these same three days Wi-Fi connected handheld devices were hit-and-miss with ftp connections whether on 2 GHz or 5 GHz.  The Wi-Fi handheld devices did connect via the AT&T LTE network however using the same software.

 

Sunday morning, 24 March, I cycled the power on the modem.  Computer connections via Ethernet were OK. Wi-Fi ftp connection worked from one device but failed on another device.  Monday morning, an Ethernet computer connected OK, but the same computer and a Wi-Fi device failed to connect to the ftp server that evening.  

 

Tuesday morning, 26 March, at 0650 MST: I tested the ftp connection from one of the Ethernet connected computers; it worked.  However at 0705 MST the same connection failed.  I don't have any way to detect packet loss but did run a traceroute from the computer that failed to connect.  It failed to complete the trace:

traceroute to www.weasner.com (206.188.193.186), 64 hops max, 72 byte packets
1 192.168.42.1 (192.168.42.1) 0.469 ms 0.232 ms 0.197 ms
2 100.67.112.193 (100.67.112.193) 0.424 ms 0.357 ms 0.256 ms
3 host1743300129217.direcway.com (174.33.217.129) 590.670 ms 566.960 ms 589.957 ms
4 host17433004221.direcway.com (174.33.221.4) 599.858 ms 637.780 ms 599.531 ms
5 ae3-651.bar2.sanfrancisco1.level3.net (4.14.107.209) 619.946 ms 597.855 ms 609.841 ms
6 * * *
7 sjo-b21-link.telia.net (80.239.195.62) 602.075 ms 627.716 ms 639.845 ms
8 defensenet-ic-302038-sjo-b21.c.telia.net (62.115.35.134) 609.946 ms 637.273 ms 589.843 ms
9 * * *
10 107.162.79.2 (107.162.79.2) 590.229 ms 596.803 ms 609.822 ms
11 * * *

 

I realize (like most technical people) that Internet routing is a mess.  But it does seem strange that the routing via the HughesNet network connection is hit-n-miss but that it always works via the AT&T LTE network.  At this point I'm not certain how to proceed.

mweasner, 

 

Are you still experiencing consistent issues with the FTP connection?

 

-Jay

My last ftp session was on Thursday, 18 April, beginning at 0725 MST.  Over the next 30 minutes I experienced ftp failures.  If I could get logged in to my server I could not get a directory listing.  If I could get a directory listing I could not upload a 10 KB text file.  It took multiple reconnections before I was successful.  So, yes, the problem still seems to be occurring.

I've kind of lost the bubble on where we were on this.

Did we ever check to see if your FTP client had a timeout setting, and if so, was it set to 120 seconds?


* Disclaimer: I am a HughesNet customer and not a HughesNet employee. All of my comments are my own and do not necessarily represent HughesNet in any way.

Both client apps I use on the desktop do have timeout settings.  They were 60 seconds on one and 30 seconds on the other.  I have set both to 120 seconds.  Will see what happens.

With the weird network delays that have been occurring, 60 would be an extreme minimum and would likely time out 10-20% of the time initially and the odds would go up exponentially over time. I'd be surprised if 30 would work if at all.


* Disclaimer: I am a HughesNet customer and not a HughesNet employee. All of my comments are my own and do not necessarily represent HughesNet in any way.

Those 60s and 30s settings worked fine until mid-December 2018.

The approximate time when the CenturyLink/Level3 backbone started going bananas.


I'm seeing other things timeout that never did before, and none of them have anything to do with HugheNet. What they do have in common is CL/L3 involved at some point in the trace. That includes oauth timeouts between my server in Pennsylvania and Twitter (2 hops from one to the other - all CL/L3), as well as between a regular web site I use and their content delivery network in Los Angeles (also only 2 hops from one to the other - all CL/L3).

 

Just so happens, looking at hop 5 in your traceroute, guess what backbone your gateway is going through?


* Disclaimer: I am a HughesNet customer and not a HughesNet employee. All of my comments are my own and do not necessarily represent HughesNet in any way.

ftp still hit-n-miss on 25 April and 26 April.

It has been several months since my last update on this issue.  The ftp protocol connections are now mostly a "miss" via HughesNet.  And to make matters worse, other protocols have become problematic.  Web access, email, and software update downloads are now experiencing significant issues.  Most of the problems appear to be with the Internet routing servers 216.6.33.133 and 107.162.79.2.  I realize that this is not a HughesNet issue.  I suppose that if HughesNet were a significant player on the Internet it might have some clout with the owners of those servers.  But I assume there is nothing HughesNet can do to improve its service to its customers when Internet routing issues like this continue to occur for months.  So I have to live with the deteriorating service.  At some point, it may become unusable.