[mythtv-users] Taking a Load Off DataDirect ???

[Brian Wood]

On Jun 11, 2006, at 9:49 PM, Robert Johnston wrote:

> On 6/11/06, Brian Wood <beww at beww.org> wrote:
>> I suspect a lot of us run multiple Myth systems. No real reason for
>> more than one in our home, but perhaps for friends, family,  
>> enemies etc.
>>
>> I suspect that in many if not most of these cases the channel lineups
>> are the same.
>>
>> How about a way to share the schedule data with multiple machines? My
>> Mom's or Brother's machine could be pointed at mine to obtain data,
>> taking some load off DataDirect.
>>
>> In major markets there might be dozens if not hundreds of Myth
>> systems, maybe a "local repeater" could service those machines?
>>
>> Or maybe Datadirect would not like that, losing the information they
>> obtain from their survey?
>>
>> I guess it depends on how much of a problem the server load is for
>> DD, and I can't answer that of course.
>>
>> But it's just an idea...
>
> Why not use one machine as a Squid proxy, and channel all the requests
> through that (As the DataDirect stuff is done over http). Then the
> subsequent requests will be served out of the Squid cache, rather than
> hitting DataDirect again.
>
> At least, in theory.

That's certainly one way to do it. The major factor is whether this  
"feature" is desirable. The apparent success of the "suggested next  
time" feature shows that cooperation with DD is possible and helpful,  
and that they are at least somewhat interested in load management.

Also, a "master" box could report what "slaves" (or perhaps just how  
many) it had passed data to when it connects to DD. Getting more  
complex, it might only pass data to a box that had a DD account.

But I wouldn't want to see too much complexity added to Myth for this  
sort of thing.