PDA

View Full Version : Remote stops working with 09/11 dongle



ggee
2007-09-30, 06:10 PM
I've been noticing that the remote for the MVP will ocasionally stop working. Or more correct, the MVP stops responding to button presses. Is there a log on the MVP that might be useful? If I telnet to the MVP, are there logs there too? Wondering if anyone else has seen this issue.

I'll try and grab the logs off gbpvr nec time it happens.

Thanks,
Greg

BTW, the loading bars under the mvpmc.org logo are a nice touch.

mvallevand
2007-09-30, 06:36 PM
If you want to help from the mvpmc side, the only way to get useful logs from mvpmc is to use this technique

http://forums.gbpvr.com/showpost.php?p=186682&postcount=37

To avoid telnet you can do this all through the config file too by redirecting output to a file ie

/etc/mvpmc/mvpmc startup options > /etc/mvpmc/dump &

To avoid hitting the network you can redirect it to a file, but you will end up filling memory.

Also if you add commands to set the time on the mvp, it makes the logs easier to interpret (I don't really care about the time zone). so add something like

rdate -s time.nrc.ca

If you do this you probably will find that 90% of the time GBPVR stops responding to the heartbeat command in these situations.

Martin

philly_phenom
2007-11-05, 11:06 PM
If you do this you probably will find that 90% of the time GBPVR stops responding to the heartbeat command in these situations.

Martin

Is there any known workaround for this? Or do you just have to power it off & restart? A bug that needs to be fixed in gbpvr (assuming Sub has the time) It's a shame because for most everything else, I like the mvpmc dongle over the hauppaugge dongle.

Thanks :)

mvallevand
2007-11-05, 11:42 PM
Is there any known workaround for this? Or do you just have to power it off & restart? A bug that needs to be fixed in gbpvr (assuming Sub has the time) It's a shame because for most everything else, I like the mvpmc dongle over the hauppaugge dongle.

Thanks :)

mvpmc is just waiting for a command so I can't really do anything about it. I've tried sending a second heartbeat but they just end up queuing up on the PC. I've actually seen the problem with the Hauppauge dongle too, but I think users are more willing to restart the Hauppauge dongle. I actually don't think this specific problem is more common in mvpmc but because Hauppauge might even restart automatically on failure it might not be as noticeable so the WAF is reduced a bit.

There may be bug in this logic somewhere, possibly in GBPVR, possibly in mvpmc but I know it cannot be easily diagnosed, because it's never reproducible. If the problem is in mvpmc it is probably in the linux kernel network drivers.

I must add that I haven't seen it as much lately, possibly the new pre 0.3.4 dongle reduces the risk. but I can't really do much more with the client side logs that might give a clue (but I doubt it).

Martin