[mythtv-users] Schedules Direct DataDirect replacement service testing

Bill Meek keemllib at gmail.com
Sun Oct 5 04:35:45 UTC 2014


On 10/04/2014 10:41 PM, Yeechang Lee wrote:
...
 >
>  From the above:
>
> 2014-10-04 20:21:54.620135 I [23743/23743] CoreContextfilldata.cpp:272 (GrabDDData) - Grab complete.  Actual data from 2014-10-04T03:18:32 to 2014-10-19T03:18:32 (UTC)
>
> Another indication that something goes wrong when the /etc/hosts
> method is used is that mythfilldatabase can't grab the next suggested
> download time:
>
> 2014-10-04 20:29:40.482204 I [23743/23743] CoreContext datadirect.cpp:1057 (GrabNextSuggestedTime) - DataDirect: Grabbing next suggested grabbing time

...

> 2014-10-04 20:29:41.289552 E [23743/23743] CoreContext datadirect.cpp:1091 (GrabNextSuggestedTime) - DataDirect: GrabNextSuggestedTime: Could not download

I'll have a look at the pastebin tomorrow, but
the failure to download the next suggested time
came up during alpha testing on *one* of my 2 hosts.

My 14.04 Mythbuntu box is OK, my 12.04 box fails
with the same error message you spotted.

The good news is that MythTV detects it and stores
the current time plus one day. Turns out that that's
what the new service return too (when it's working!)

If you use your favorite protocol analyzer, then I
suspect you'll see the server send a FIN in the last
packet of compressed data, however the client doesn't
return a FIN, ACK and then tries to send the request
for the next suggested time to the conversation that
was just finished.

The server, properly, sends back an RST. When run on
my 14.04 box or with the 'old' server, the FIN, ACK
is returned and all is well.

-- 
Bill


More information about the mythtv-users mailing list