Streaming movies - upgrade to 6.3.5 breaks this


cowger

Recommended Posts

Hello,

 

Complete newbie here in search of remedial assistance.  A bit of background:  I purchased my NAS box from Cinemar (home automation software) with unRAID pre-installed on it, so I don't have much knowledge at all about the system.  After running version 5.0.6 on it for years, I decided to update it to 6.3.5.  That upgrade went flawlessly (at least I thought so at the time), and the array has been up ever since (about a week).

 

However, now my movie streaming experience is broken.  DVD files will run for 30 minutes or so and then go into this "slow motion" mode, which I think I've seen others describe here as "stuttering".  It seems to mimic what an actual disc playback is like if the player encounters a scratch on the disc.  BlueRay files will exhibit this behavior in less than a minute.

 

My configuration is this:  NAS (unRAID) <-> Cisco SG200 switch <-> Dune media player.

 

I'm attaching a screenshot of the "Main" view, along with my debug file, in case that helps.

 

Other things I've tried:

1) I installed the "Fix Common Problems" plugin and ran that.  The only error was that I didn't have "Plugin Update Check" enabled.

2) I installed the  "Tips and Tweaks" plugin and ran that.  Unfortunately I don't have the first clue as to what to tweak as a result of that.

 

I did search through the forum here for about an hour and found others with seemingly the same issue after upgrade to version 6.xxx.  Unfortunately I didn't find any answers that seemed to explain what might be the difference between the older version of unRAID (that ran just fine for years) and version 6.3.5 regarding streaming media.  I've also seen users report that they are using Dune with 6.3.5 and it works just fine, so I'm sure it's something I'm doing wrong.

 

I am now unable to view my movie library so any ideas on what I can try are very welcome.

 

Bryan

 

Main screenshot.JPG

mldata-diagnostics-20171016-0726.zip

Link to comment

Hi -

 

Do you have the specs on your server?  I.e., CPU, memory, etc?  (Looks like a Core i3 with 8GB but want to confirm.)  Also, what does your media server/player environment look like?  You mentioned Dune, so are you using Dune players to access the media files directly on your server?  Or is anything like Plex involved?  It looks like you setup Docker... 

 

FYI, you have a bunch of really full ReiserFS disks - that's not good.  It's a deprecated file system that's know to behave poorly when full... You also have something going on with ntp.

Link to comment
14 minutes ago, tdallen said:

Hi -

 

Do you have the specs on your server?  I.e., CPU, memory, etc?  (Looks like a Core i3 with 8GB but want to confirm.)  Also, what does your media server/player environment look like?  You mentioned Dune, so are you using Dune players to access the media files directly on your server?  Or is anything like Plex involved?  It looks like you setup Docker... 

 

FYI, you have a bunch of really full ReiserFS disks - that's not good.  It's a deprecated file system that's know to behave poorly when full... You also have something going on with ntp.

 

First off, thank you, tdallen, for taking the time to ask some questions.  

 

I'm attaching a screenshot from the "System Profiler" plugin, which confirms your assumption of a Core i3 processor with 8GB of memory.

 

My media player is just a single, completely stock Dune player connected via a 100Mb link, one hop to the switch and one hop to this NAS server.  I don't know what Plex is, so I can't answer that question.  I also don't know what Docker is, sorry.

 

And yes, you're correct that my smaller disks are very full (screenshot of that attached, too).  I've just been adding more disks as the previous one filled up, assuming that anything less than 100% is okay.  Apparently that's not the case and I should better balance their contents.   I will need to go read up on how to do this.  Should I assume that they should be below 90%?  80%?

 

Thank you,

Bryan

 

system info.JPG

disk stats.JPG

Link to comment
27 minutes ago, ashman70 said:

There is a plugin called unbalance that would allow you to manually move files around the free up space, you could try that and see if it alleviates the problem.

 

Thanks, that looks like what I need!  Unfortunately, once I've enabled it, I click on "Open Web UI (v3.3.3 )" and it just pulls up a blank web page.  I'll keep trying...

 

UPDATE:  I just tried opening unbalance via Chrome instead of Firefox.  Now it's working...  : )

Edited by cowger
Link to comment

I appears that you are spinning your disk down when they are inactive.  What is  your Default spin down delay:  set to?  Since the problem seems to start about thirty minutes after you start watching, I would recommend you try 45 minutes or an hour.  (I have sometimes observed if the playback application/device stops for 'lack of data', the condition you describe will occur.)

Link to comment
9 minutes ago, Frank1940 said:

I appears that you are spinning your disk down when they are inactive.  What is  your Default spin down delay:  set to?  Since the problem seems to start about thirty minutes after you start watching, I would recommend you try 45 minutes or an hour.  (I have sometimes observed if the playback application/device stops for 'lack of data', the condition you describe will occur.)

 

Thanks, Frank, that's something to try (set them to never spin down and then try watching a movie).  But since a Blueray file will start this behavior virtually right away, I'm not sure this is the culprit...

Link to comment

Here's an update, along with a few more (basic, possibly stupid) questions:

 

I've spent some time with the "unbalance" plugin to reallocate files from drives 1-5.  I also took the opportunity to cull out some of the movies we'll never watch again.  With those two things I now have a slightly better-looking Disk Usage chart (below).  Not knowing a thing about the weakness(es) of the ReiserFS, do I need to get them all below 80%?  70%?

 

Other questions:  Why is there still 40+ GB of data on the Cache drive?  Isn't mover's job to move that to other disks overnight?  I also tried manually invoking mover this morning with no apparent effect.

 

A quick test of streaming a 1080p movie this morning resulted in no apparent failures (only watched about 5 minutes, though).  This same movie failed previously (went into stutter mode) before the stars in the Paramount preamble had even stopped moving.

 

One other data point that may be relevant:  My 2nd use of this NAS box / Ethernet switch is to manage traffic from 7 IP security cameras.  Could that be part of the problem?  Cisco claims that this particular switch has a total bandwidth capability of 52 Gbps, or something north of 5 GB/s.  Would a bunch of security camera traffic possibly cause a problem?  Assuming each is saturating a 10Mb link, that should still be less than 6 or 7 MB/s total, right?  Files that are saved by the security app are written first to the server (running the security app), and then eventually moved to the NAS box after 7 days.

 

 

 

disk stats 17Oct.JPG

Edited by cowger
Link to comment

I know you mentioned that you aren't familiar with Docker, but it looked from the logs like it might be setup (and would explain the use of the cache drive).  Click on the Docker tab and post that screenshot.

 

Docker is the main thing that has turned unRAID from just a NAS to an application platform.  There are now dozens and dozens of apps that run on unRAID as Docker containers.  To support this, the cache drive has become the defacto "application drive" in unRAID 6+ and applications and their data are typically permanently stored on the cache drive.

Link to comment
52 minutes ago, tdallen said:

I know you mentioned that you aren't familiar with Docker, but it looked from the logs like it might be setup (and would explain the use of the cache drive).  Click on the Docker tab and post that screenshot.

 

Docker is the main thing that has turned unRAID from just a NAS to an application platform.  There are now dozens and dozens of apps that run on unRAID as Docker containers.  To support this, the cache drive has become the defacto "application drive" in unRAID 6+ and applications and their data are typically permanently stored on the cache drive.

 

Here's the screenshot of the Docker tab.  Is there anything here that explains the contents on the cache drive?

 

docker tab.JPG

Link to comment
5 minutes ago, tdallen said:

Nope.  Time to explore the cache drive and see what's there...

 

Yes, good idea.  What I found is that the 40+GB  of data is all the saved videos from my security cameras.  After spending a week on the security camera server, they are written (by the security camera app) to this location:  \\MLDATA\Archives\security   (Then after aging there for 28 days, they get deleted, again by the security cam app.)

 

And below is a screenshot of the "Archives" share.  

 

Questions:

1) Why do these files get placed onto the cache drive, when "Use cache disk:" is set to "No"?

2) Why doesn't mover push these to the main drives at 3:40am?

 

archives share.JPG

Link to comment
6 minutes ago, cowger said:

 

Yes, good idea.  What I found is that the 40+GB  of data is all the saved videos from my security cameras.  After spending a week on the security camera server, they are written (by the security camera app) to this location:  \\MLDATA\Archives\security   (Then after aging there for 28 days, they get deleted, again by the security cam app.)

 

And below is a screenshot of the "Archives" share.  

 

Questions:

1) Why do these files get placed onto the cache drive, when "Use cache disk:" is set to "No"?

2) Why doesn't mover push these to the main drives at 3:40am?

 

archives share.JPG

Not sure why 1) is happening.

2) is explained by the fact that with Use Cache=No set then if any files for that share end up on the cache drive for some reason they are ignored by Mover.    Mover will only move files from cache to array for shares where use Cache=Yes.

Link to comment
11 minutes ago, tdallen said:

I would imagine that something is writing directly to /mnt/cache/Archives rather than /mnt/user/Archives ?

 

What is your security camera app and where does it run, how is it configured, etc?

 

The security cam app is called BlueIris.  Below is a shot of the configuration that tells it where to move clips after 7 days.

 

BlueIris runs on a standard dual-socket Intel server, in this case called "MLServer".  unRAID is running on a separate storage server called "MLData".  (Cinemar, the original vendor of this, is the one who named them; ML stands for "MainLobby", which is their movie automation app that also runs on MLServer.)

 

My format of \\MLDATA\Archives\security is slightly different than yours, but it's not obvious to me how/why that is writing to cache.  In any case, I've kind of highjacked my own thread here on this cache disk content thing... 

 

The main issue (movies not playing) remains the central issue here.  I owe you an update on whether I can reliably watch a movie from start to finish.  I'll work on that tonight.  Perhaps it was simply that I had too little spare capacity on disks 1-5, and your original statement that ReiserFS runs poorly in this situation was the root cause.  

 

Once I get this working, then I'll perhaps work on migrating to XFS.  (While a noob, I do realize that there's a process and you can't simply start reformatting drives.)

 

Thanks for everything so far!

Bryan

 

blueiris.JPG

Link to comment

Ok, that helps.  Then under the Shares tab you should have an "Archive" share which if you click into will show as "Use cache disk: only".  If that is the case then Cinemar set this up to use only the cache disk for Blue Iris files - never allow mover to copy them to the array - not sure why.

Link to comment
Just now, tdallen said:

Ok, that helps.  Then under the Shares tab you should have an "Archive" share which if you click into will show as "Use cache disk: only".  If that is the case then Cinemar set this up to use only the cache disk for Blue Iris files - never allow mover to copy them to the array - not sure why.

 

I posted a pictures of the Shares tab for "Archives" a few posts up.  It's set to "Use Cache disk: No".  

 

Remains a mystery...

Link to comment
46 minutes ago, tdallen said:

Odd.  Are the files on the cache drive current, i.e. within the last 28 days?

 

Yes, date ranges from 10/3 to 10/7. (I believe I had wiped the primary repository (on MLServer) clean back on 10/3 when I started this whole upgrade adventure...)

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.