Extender issues after start TV recording playback

[Fuzzy]

Originally Posted by SomeWhatLost

fwiw, I figured out what was causing this for me, in my case it was purely a SATA/SAS bandwidth issue.
I moved my recording drives off the SAS backplane and connected them straight to the motherboard and the weird pausing/fast forwarding/skipping thingy/non-responsiveness has completely gone away...

only issue I have left is the spinning circle of doom while scrolling through my TV shows... that is still a bit odd, will be going fine, that just sit there for a while, then all the show banners fill in and it comes back to life... it is almost like gemstone is waiting to load the show banners before allowing things to move on... which is odd because gemstone never seemed to care with V7...
even more odd is once you go through waiting for the circle of doom to do its thing, it is done... no more circles until you shut down the extender for the night and come back tomorrow.... like it loaded up a cache or something and then loses it over night...

That sounds like it's waiting for a drive to spin up where the fanart is stored, perhaps. Where are your apps & appdata shares stored, and what is the cache policy, and the folder splitting policy? (I think I remember you saying you were using unRAID).

[Taddeusz]

Originally Posted by SomeWhatLost

fwiw, I figured out what was causing this for me, in my case it was purely a SATA/SAS bandwidth issue.
I moved my recording drives off the SAS backplane and connected them straight to the motherboard and the weird pausing/fast forwarding/skipping thingy/non-responsiveness has completely gone away...

Really, what SAS adapter do you have? I have an IBM m1015 cross-flashed to an LSI 9211 and have had weird issues with sudden jumps during smooth FF & REW in both Windows and now unRAID.

[EnterNoEscape]

Originally Posted by SomeWhatLost

fwiw, I figured out what was causing this for me, in my case it was purely a SATA/SAS bandwidth issue.
I moved my recording drives off the SAS backplane and connected them straight to the motherboard and the weird pausing/fast forwarding/skipping thingy/non-responsiveness has completely gone away...

only issue I have left is the spinning circle of doom while scrolling through my TV shows... that is still a bit odd, will be going fine, that just sit there for a while, then all the show banners fill in and it comes back to life... it is almost like gemstone is waiting to load the show banners before allowing things to move on... which is odd because gemstone never seemed to care with V7...
even more odd is once you go through waiting for the circle of doom to do its thing, it is done... no more circles until you shut down the extender for the night and come back tomorrow.... like it loaded up a cache or something and then loses it over night...

A while back when I used unraid, SageTV would often show that circle while drives were spinning up. If you allow your drives to spin-down, this will always be something you'll deal with. I started moving back to unraid this past weekend and one of the things I did was create a share that only goes to one of the 3TB hard drives for recordings. I then excluded this same disk from any other shares. This disk is also using the cache. This means that I will only need one hard drive to spin up for recordings. It's the only thing I wanted to be sure would be fairly responsive. It can always be expanded if needed, but 3TB is a lot of space for "temporary."

I don't access my archived TV and movies too often, so I care a little less about how those end up being laid out, but I did for example ensure that one TV show only spans one disk.

You can also do spin up groups so that more drives will spin up at the same time which might reduce the amount of time you're staring at that circle.

Edit: Also take a look at the Folder Caching plugin for unRAID. If you spin your drives down, this can often save you from spinning one up just because the folder is being accessed.

[starfire]

Hi All,

Thanks for the replies.

The key thing is that apart from moving from a Windows 10 VM on unRAID to the docker everything else has remained the same in terms of drives & media locations, I record to my SSD cache drive as before and the mover kicks at night and transfers recordings to the array. No drive locations have changed.

I can experience the delay on new recordings that are sitting on the continually spinning SSD cache!

Really has me baffled as I have NO spinning circles, purely it's a failure to respond to skip forward or back media play - the commands seem to be buffered for a period of x seconds then suddenly kick in.

As I seem to be the only one suffering from it then I guess something in my config is causing it, I'll try to narrow it down with further testing.

Cheers.

[stuckless]

Originally Posted by starfire

Hi All,

Thanks for the replies.

The key thing is that apart from moving from a Windows 10 VM on unRAID to the docker everything else has remained the same in terms of drives & media locations, I record to my SSD cache drive as before and the mover kicks at night and transfers recordings to the array. No drive locations have changed.

I can experience the delay on new recordings that are sitting on the continually spinning SSD cache!

Really has me baffled as I have NO spinning circles, purely it's a failure to respond to skip forward or back media play - the commands seem to be buffered for a period of x seconds then suddenly kick in.

As I seem to be the only one suffering from it then I guess something in my config is causing it, I'll try to narrow it down with further testing.

Cheers.

You might have mentioned, but what is your client? (Android MiniClient, HD300, PC, etc) and have you tried different clients... Do any other operations feel buffered... ie, when you are navigating menus, etc.?

[SomeWhatLost]

Originally Posted by Taddeusz

Really, what SAS adapter do you have? I have an IBM m1015 cross-flashed to an LSI 9211 and have had weird issues with sudden jumps during smooth FF & REW in both Windows and now unRAID.

oddly enough, it is an actual LSI 9211-8i...

[SomeWhatLost]

Originally Posted by Fuzzy

That sounds like it's waiting for a drive to spin up where the fanart is stored, perhaps. Where are your apps & appdata shares stored, and what is the cache policy, and the folder splitting policy? (I think I remember you saying you were using unRAID).

fanart/apps/appdata are all on a micron M500 SSD, spin up time should be fairly quick I would hope?
no folder splitting or anything as it is an "Unassigned Drive" outside of the array...
yes on unRAID... I liked what they did with old v4, but I really like the new v6... such a perfect setup for the home.

[Taddeusz]

Originally Posted by SomeWhatLost

oddly enough, it is an actual LSI 9211-8i...

Interesting. The whole reason I got the m1015 card was that I was having performance problems with live comskip using the integrated SATA controller. At the time I had an Intel G35 based board and Core 2 Quad CPU. Putting all my drives on the SAS controller fixed those performance problems. I actually don't remember when FF/REW jumping started. I think it's been a problem all along. I assumed it had something to do with the recordings.

[starfire]

Originally Posted by stuckless

You might have mentioned, but what is your client? (Android MiniClient, HD300, PC, etc) and have you tried different clients... Do any other operations feel buffered... ie, when you are navigating menus, etc.?

Only noticed on my 2 x HD300s, only other client I use is my Windows PC and that never has the issue.

Other operations seem fine, you can go around the menu's etc. fine, it's just when you initially select a recording for playback.

My only thought is that perhaps for some reason Sage under Linux wants access to some shares that are spun down on my unRAID server - BUT, I'd doubt the code for playback is different at this point under Linux than Windows for an extender?....


edit: Guess I could totally disable all spin down for a few days, that would prove if that's the cause

[trk2]

Originally Posted by SomeWhatLost

fanart/apps/appdata are all on a micron M500 SSD, spin up time should be fairly quick I would hope?
no folder splitting or anything as it is an "Unassigned Drive" outside of the array...
yes on unRAID... I liked what they did with old v4, but I really like the new v6... such a perfect setup for the home.

So this sounds similar to a problem I used to experience and it was related to sleeping hard drives.. I solved it by setting the Sage media scan time to less than my hard drive sleep time. So Sage keeps all the drives awake whenever it is in active use.

[SomeWhatLost]

Originally Posted by SomeWhatLost

fanart/apps/appdata are all on a micron M500 SSD, spin up time should be fairly quick I would hope?
no folder splitting or anything as it is an "Unassigned Drive" outside of the array...
yes on unRAID... I liked what they did with old v4, but I really like the new v6... such a perfect setup for the home.

just a FWIW type deal...
well, I think I found my issue with the spiny circle of DOOM!!!
seems my Cache SSD's were dying... (stupid OCZ Revo drive) as far as I know, the only sage related thing there was the docker.img file... not sure how that relates to sage, but anyway, new micron 500 series SSD to replace the OCZ, new docker.img, and everything seems happy...