[VU+ Solo2] Help with wierd error messages

[bbbuk]

I did a fresh flash with 20 again but this time I didn't get any update error message but I did when I first flashed 20.

Nothing has changed with regards to my network, etc so I doubt it's related to my network.

This must be a right pain to track down with it being so intermittent

 

[birdman]

Sorry for repeating things but thinking out loud - assuming the server end is ok it's not a speed / duplex mismatch that's appeared in certain NIC drivers is it?l

The NIC driver is in the kernel, and mine hasn't changed since I got the box a year ago.

If auto-negotiation had failed I'd notice it when connecting to the box from my PC and I'd be very surprised: I've never seen a Linux box get auto-negotiaion wrong once the router/switch has been configured to actually allow (rather then be hard-wired so that a Solaris system would work and some operator had put cables back into the wrong sockets.)

 

[birdman]

This must be a right pain to track down with it being so intermittent

There seem to be two issues.

• The intermittent network timeouts (at the moment it's running fine).
• The Enigma2 coding which turns these timeouts into crashed, rather than just reporting a failure.

 

[kill_bill]

Day 3 of fresh install with background update check off and all good. As mentioned before, nothing has changed on my network for the past 18 months

 

[judge]

Day 3 of fresh install with background update check off and all good. As mentioned before, nothing has changed on my network for the past 18 months

And no fix has gone into released images yet...

 

[kill_bill]

Ok, think I've got it narrowed down. Turned off DHCP tonight and manually changed the IP address. Hour later it crashed. Didn't save the logs on the USB for some reason.

Flashed back a full image backup I took two days ago with dhcp on and all stable

 

[birdman]

I never use DHCP for static system at home (apart form laptops - which get fixed IP addresses in the DHCP server).
And I saw the problem.

It seems to be related timeouts to the update server.
A patch in that area did get into the git repository yesterday.

 

[kill_bill]

Since my original blip had been solid for past two days with DHCP enabled and background checks off. Will leave as is :thumbsup:

 

[emmaian]

I have to also noticed that if dchp is off so giving static ip it crashes, so i though ok lets turn dchp on but this time set Vu duo2 to use the same ip via the modem, again the same crashes, turn off the fixed ip on modem and once a reflash is done all good runs sweet. My conclusion is it is network related and also down to fixed ip either done by box or fixed in modem.Hope this helps anyone.

 

[abu baniaz]

This issue should be fixed in next image (022). My receiver has been checking the update server every five minutes with a modified file for the last week, no crash.

Reminder...Do not update to 022. install it and restore settings.

 

[kill_bill]

So there was a problem deep down which wasn't our faults, good to know after the initial finger pointing :thumbsup:

 

[Sicilian]

So there was a problem deep down which wasn't our faults, good to know after the initial finger pointing :thumbsup:

What finger pointing?? There's two issues in this thread. One is online update check when users receiver looses connection and the other a phantom unexplained crash that results in needing a re-flash.

The update check users receivers are loosing network connection, hence the error message.


Sent from my iPhone using Tapatalk

 

[birdman]

This issue should be fixed in next image (022). My receiver has been checking the update server every five minutes with a modified file for the last week, no crash.

But has it ever seen, and handled gracefully, a socket timeout?
It's always been OK if it gets through, or fails to connect at all.

 

[abu baniaz]

Had a couple of "There has been an error.......". Didn't crash, receiver continued to work.

Would I be correct in assuming you have the two modified files on your system and you still have an issue?

Sent from my HTC One M9PLUS using Forum Fiend v1.3.3.

 

[birdman]

Would I be correct in assuming you have the two modified files on your system and you still have an issue?

If that is aimed at me then no, I don't have the changes I spotted in git. I just switched off the auto-check and left it at that.
I was going to look at how to trap a socket timeout in python, but I wasn't able to produce one to test. I could get it to fail to connect, or wait "forever" but not timeout - mind you, I didn't try all that hard.

 

[Davitat]

If that is aimed at me then no, I don't have the changes I spotted in git. I just switched off the auto-check and left it at that.
I was going to look at how to trap a socket timeout in python, but I wasn't able to produce one to test. I could get it to fail to connect, or wait "forever" but not timeout - mind you, I didn't try all that hard.

It seems at least partially the same problem:
http://www.world-of-satellite.com/showthread.php?48395-VU-Solo-2-VIX-image-Hades-VIX-3-2-Crashing

1) A network software related bug (driver?)
2) An error handling bug.

Seems also not OpenVIX specific as I got the same errors with BlackHole and OpenPLI.

 

[bbbuk]

Just tried to run "opkg update" and first two feed files (3rd-party and all packages) got an "wget: bad address 'www.openvix.co.uk' " error but rest of the feed files updated fine.

I then re-ran the same command straight after and it updated all the feed files fine.

It would appear to be an intermittent problem with openvix.co.uk.

Could this be related to the recent re-design (ie was a server change done or something else done at same time maybe)?