[mythtv-users] Some master/slave backend advise required

Tue Jun 3 05:35:40 UTC 2014

On 03 Jun 2014, at 05:19, Henk D. Schoneveld <belcampo at zonnet.nl> wrote:

> 
> On 02 Jun 2014, at 20:59, Johan Van der Kolk <johan.vanderkolk at gmail.com> wrote:
> 
>> Hi,
>> 
>> I’m running myth 0.27 (v0.27.1-7-g41d04b6) on Ubuntu server 14.04 with an i5(3GHz) and 4GB memory. I’m using 4 DVB-S2 tuners as source.
>> Due to relocation of the dish I have to install a slave backend. I can’t (don’t want to) move my server to the garden shed.
>> 
>> I have found that recording about 8-10 HD channels simultaneously will start killing processes on the backend (not the myth processes though), but the CPU load of the DVB-s2 stuff by itself requires 100% cpu. Mythcommflagging did not help either, although I could limit the simultaneous jobs to get more breathing space. 
>> So my thought was to solve both problems at the same time with a new slave backend, where the dvb-s2 stuff is running, by itself. Storage and commflagging still to be done on the master backend (which has 8TB of ZFS storage)
>> 
>> I can see two issues now:
>> Bandwidth: When recording 10 channels (what I want to achieve), I estimate worst case (based on what i recorded so far) 3GB per channel per hour, or 833Mb/s one way traffic, without other overhead. And not watching anything…
> 10 * 3GB = 30.000MB per hour. 30.000 / (60 * 60)sec = 8,33MB/s
> Every single current HD should be able to do that if it doesn’t write randomly, in small blocks. I use xfs, where it is possible to allocate blocksizes of 1M - 500M to solve fragmentation of recordings. see:  http://www.mythtv.org/wiki/Optimizing_Performance 

For those who understand bonnie ( i don’t). This is a command line taken from https://calomel.org/zfs_raid_speed_capacity.html.
I ran it on my zfspool.

Version  1.97       ------Sequential Output------ --Sequential Input- --Random-
Concurrency   1     -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
Zolder-server   80G           318688  37 136544  25           269337  20  38.6   2
Latency                        4301ms    1176ms               549ms     658ms

>> How to solve this, and does myth traffic between master and slave benefit from Jumbo Frames. I could use two network cards in each machine, and create a 2 x1 Gb trunk between the switches.
>> Second part of the problem might be that the ZFS file storage is not fast enough. (upgrade to SSD maybe)
>> 
>> Configuration:
>> Is it possible to configure mythtv in such a way that it does what I want, or are there better ways to do this? 
>> 
>> Any help appreciated!
>> 
>> Johan
>> _______________________________________________
>> mythtv-users mailing list
>> mythtv-users at mythtv.org
>> http://www.mythtv.org/mailman/listinfo/mythtv-users
>> http://wiki.mythtv.org/Mailing_List_etiquette
>> MythTV Forums: https://forum.mythtv.org
> 
> _______________________________________________
> mythtv-users mailing list
> mythtv-users at mythtv.org
> http://www.mythtv.org/mailman/listinfo/mythtv-users
> http://wiki.mythtv.org/Mailing_List_etiquette
> MythTV Forums: https://forum.mythtv.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mythtv.org/pipermail/mythtv-users/attachments/20140603/6a2a42e8/attachment.html>