<br><div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>From: Martin Turner <<a href="mailto:xmode@westnet.com.au">xmode@westnet.com.au
</a>><br>Subject: [mythtv-users] Unusual problem<br>To: Discussion about mythtv <<a href="mailto:mythtv-users@mythtv.org">mythtv-users@mythtv.org</a>><br>Message-ID: <<a href="mailto:46DD33FE.2090205@westnet.com.au">
46DD33FE.2090205@westnet.com.au</a>><br>Content-Type: text/plain; charset=ISO-8859-1; format=flowed<br><br>Hi MythList,<br><br>This isn't strictly a Myth question, but only my master Myth backend is<br>doing this so I thought I'd ask here. The box its self has been running
<br>more or less fine for about 2 years now except that recently (the last 2<br>weeks) it has developed a stutter.. Now I say stutter because it<br>sounds/looks like it gets stuck in a second loop for about 4-6 seconds<br>
then continues on fine like nothing ever happened. I don't have anything<br>in the logs (myth or syslog) to indicate what is going on and im<br>basically stuck.. As you can imagining its extremely annoying having it<br>
do this at random..<br><br>This is what I have observed/done so far..<br>- Seems to be at random. Doesn't coincide with any regular running task<br>or recordings in myth. Can do it while myth is recording or not.<br>- Nothing is logged. If myth is recording I get a log about "
<br>cx88_wakeup: 31 buffers handled (should be 1)" but after investigation<br>im putting that down to the cards internal buffers filling up during the<br>stutter then all being read at once rather than it being the error thats
<br>causing this.<br>- New kernel didn't fix the fault (gentoo system, roll my own kernel)<br>neither did changing the kernels tick timer or making it a tickless<br>kernel. Also, the problem didn't start after a kernel change.
<br>- Seems to have started after the 0.20.2 upgrade but that may just be<br>coincidence. Is it safe to roll back to test?<br>- The clock in the system actually stops for the time its stuttering and<br>needs to re-sync. NTP fixes this fairly quickly, but interesting..
<br><br>Any help would be great, I'm happy to post more info about my system if<br>its needed.</blockquote><div><br>My first guess would be that it's a disk seek issue. Use your distros tools to look into IO response times and possibly disk fragmentation. Try defragging even if it's only slightly fragged. If it persists with no explanation, try replacing the drive with a newer, higher speed drive with a larger cache. I have done so in the past using dd to image it over. Failing drives often leave no log evidence of their impending doom.
<br></div><br></div><br>