<div dir="ltr"><div><div>I doubt it's a RAID 5 performance issue. The performance of the disk array for both reads and writes is pretty good. Write performance is around 320Mbps ...<br><br># dd if=/dev/zero of=/export/test2.img bs=1G count=1 oflag=dsync<br>1+0 records in<br>1+0 records out<br>1073741824 bytes (1.1 GB) copied, 22.4376 s, 47.9 MB/s<br><br>and read performance is almost 800Mbps ...<br><br># dd if=/export/record/1752_20160312073000.mpg of=/dev/null bs=8k<br>2027724+1 records in<br>2027724+1 records out<br>16611117128 bytes (17 GB) copied, 171.084 s, 97.1 MB/s<br><br><br></div>So back to my original question, are those RingBuf errors indicating that the myth backend stopped receiving data from the HD Homerun Prime or the myth backend was unable to read data from the file and stream it to the frontend?<br><br></div>Thanks for the help ...<br><div><div><br><br></div></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Mar 21, 2016 at 6:55 PM, Mike Thomas <span dir="ltr"><<a href="mailto:mt3@pfw.demon.co.uk" target="_blank">mt3@pfw.demon.co.uk</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Mon, 21 Mar 2016 16:45:06 -0700<br>
Gabriel Kuri <<a href="mailto:gkuri@ieee.org">gkuri@ieee.org</a>> wrote:<br>
> While watching TV, the video on the remote frontend sometimes freezes<br>
> or pixelates. This last time, I had to exit back to the main menu and<br>
> go back in to the channel to continue watching because it just froze<br>
> for way too long. Later, I looked at the logs on the backend and<br>
> noticed this in the log about the same time it froze ...<br>
</span>> [snip]<br>
<span class="">> The video is on a RAID 5 array that is healthy and the SMART status<br>
> of all 4 drives is good. So I don't believe I have a failing drive.<br>
<br>
</span>Dear Gabriel,<br>
<br>
I will let others remark on the specific messages but I will respond to<br>
the RAID 5 point.<br>
<br>
RAID 5 is the slowest form of RAID for writes. To update a single disc<br>
block (when it's not in cache) on a four disc RAID 5 array involves<br>
four disc reads, a computation which can only start after all four<br>
reads have completed, and then two disc writes. The writes will have<br>
missed their time slots and thus must wait for the platters to rotate<br>
into position.<br>
<br>
Although single block I/O is not the bulk of what's going on on a myth<br>
system (video files are big), inode updates are, and they are updated<br>
very frequently.<br>
<br>
As to a failed disc in a raid array, no you don't have one. You would<br>
*really* know about it then, believe me.<br>
<br>
The message from me is if you want to use RAID for data security, use<br>
RAID 1. RAID 5 is fine for archive storage but not for active<br>
filesystems. *If* you have a performance problem and *if* someone thinks<br>
it is disc-related, I would point the finger at that. This comes up on<br>
the list from time to time.<br>
<br>
Mike.<br>
_______________________________________________<br>
mythtv-users mailing list<br>
<a href="mailto:mythtv-users@mythtv.org">mythtv-users@mythtv.org</a><br>
<a href="http://lists.mythtv.org/mailman/listinfo/mythtv-users" rel="noreferrer" target="_blank">http://lists.mythtv.org/mailman/listinfo/mythtv-users</a><br>
<a href="http://wiki.mythtv.org/Mailing_List_etiquette" rel="noreferrer" target="_blank">http://wiki.mythtv.org/Mailing_List_etiquette</a><br>
MythTV Forums: <a href="https://forum.mythtv.org" rel="noreferrer" target="_blank">https://forum.mythtv.org</a><br>
</blockquote></div><br></div>