jhaar-ourshack-com at whanau
Jun 19, 2012, 4:59 PM
Post #1 of 7
Totally weird situation here. Over the weekend I replaced my backend
server with a CentOS-6/x64 system running
mythtv-backend-0.25-286.el6.x86_64, with a USB DVT device. Worked fine -
to begin with. Yesterday I noticed mythweb was no longer working - long
story short - mythbackend is now accepting TCP connections on it's
management ports - but isn't responding.
I went as far as backing up the mysql database and then dropping it and
reinitializing using mythtv-setup. That all worked fine but mythbackend
is still broken. I removed ipv6, the firewall, the USB card, rebooted:
mythbackend logs showed it was happy - but it still didn't respond to
So far so normal - here's the kicker. "lsof -ni" shows mythbackend is
LISTEN-ing on all the correct ports, telneting to port 6544 *succeeds*,
"lsof -ni" shows that ESTABLISHed connection - but strace-ing the
mythbackend process *does not show it received the connection*...
It's as if the OS did the three-way TCP handshake and then failed to
pass control to the application??? If I strace xinetd for instance and
then connect to my rsync server, strace sees the incoming connection and
the read/writes that occur - mythbackend does not... The logs do show it
attaching to all the interfaces (ie the logs agree with lsof -ni)
Brand new hardware, RAID1 disks - what the hell!?!? "rpm -V
mythtv-backend" shows the binaries are not corrupt too.
Any ideas - i've run mythbackend with "-v debug" and it shows no issues
- weelllll, the logs do show Bonjour is broken (and it shouldn't - it
was working in the weekend after I figured out to add the 239/8 route) -
but I doubt that's related.
mythtvnz mailing list
mythtvnz [at] lists