Playback Pauses/Hangs then FF catchup

[dlandrum]

I have recorded to a RAID 5 array for some time with 6 tuners.

I switch from a high-speed 5-drive array to a low-speed 4-drive array.
That is when my issue started.

I went to slower speed drives as they were more green; low heat, energy usage, noise.


Is it possible to record all shows to one drive(SSD), and have Sage or a plugin copy the show to the RAID drive?

[Spectrum]

SJQ can move recordings around for you, but I don't know what recording everything to a single SSD would do for the lifespan of the SSD. I would think there would be enough writes involved to seriously shorten its usable lifetime.

The other option would be to take the drives out of a RAID config and have 4 volumes for recording. With V7 there is a setting to optimize recording for performance so it Sage would balance out the recordings across the spindles. Upside is better performance, downside is losing the ability to recover from a single drive failure.

[ldw58]

I have been noticing the problem for the past week.

1. It happens for aboug 2 1/2 minutes at a time.
2. when I RDP into WHS & run Windows Task Manager I see that there are four individual new FFMPEG processes running and that they are killing my system performance. The problem goes away (the 2 1/2 minutes are up) when all the FFMPEG processes go away.
3. I cannot recall it ever happening when I was watching a video that hadn't already been transcoded (I use the "Universal" setting on MediaShrink). All my transcoded videos go into the pool. I record live video onto a dedicated caching HDD and then move it to the pool once they've been transcoded.
4. It does not happen to BluRay or DVDs.

UPDATE ON 12/10/2010:
I've tracked my problem down to MediaShrink/HandbrakeCLI execution (although it is not clear from the logs what is going wrong). After a HandbrakeCLI crash subsequent executions of MediaShrink occasionally just spawn up to five FFMPEG processes and bring the WHS system to a crawl. The programs being recorded by Sage during one of these crashes have about ten minutes of corrupt data within them.