Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[glynor]

The copy of MC I use at the office for media management of my streaming servers has been acting troublesome of late, and I'm not sure why or if there is anything I can do to fix it.

What is happening is this:

Periodically, and seemingly at random, MC's UI completely freezes.  This manifests as a spinning circle mouse pointer, and the UI going "grey" in Windows, with Windows popping up the "Application is busy, terminate or wait?" dialog.  This happens, seemingly "periodically" and when it "unfreezes" then MC goes back to being normally responsive.  But within a minute or two, usually, it will come back.

I'm not sure how to fix it.

Here's a log from it happening recently:
http://glynor.com/files/jriver/JRiver_Log-2013-02-20_11-28-36.zip

In that log, I opened MC, left it running for a bit, and then tried to just open the Options panel.  When I went up and clicked on the Tools menu item, it froze for several seconds (more than 10), and gave me the "grayed out UI" and spinning mouse cursor.  I waited until it "unstuck" and then it came back to life and I immediately made a log.

I've tried this:

Options > Tree & View > Advanced > Display missing file images in lists: Disabled
Podcast Auto-Downloading is Disabled
Auto-Import:
Two UNC paths monitored.  They're fairly large, complex (and somewhat slow) SMB shares on our network.
However, I've disabled:
Analyze Audio
Build Thumbnails
Get Cover Art
Get Move & TV Info
Ignore files previously removed (I want everything on the drives in the Library)
Update for external changes

The only stuff "turned on" in there is Fix Broken Links and the Write file tags setting.

The Library itself is on C: (which is the VM's virtual disk, but that itself lives on a fast SSD), so it shouldn't be a problem accessing the Library.  My JRMark from within the VM looks like this:

Code: [Select]

=== Running Benchmarks (please do not interrupt) ===

Running 'Math' benchmark...
    Single-threaded integer math... 5.699 seconds
    Single-threaded floating point math... 3.195 seconds
    Multi-threaded integer math... 2.841 seconds
    Multi-threaded mixed math... 1.624 seconds
Score: 1422

Running 'Image' benchmark...
    Image creation / destruction... 0.781 seconds
    Flood filling... 1.739 seconds
    Direct copying... 0.731 seconds
    Small renders... 1.931 seconds
    Bilinear rendering... 1.397 seconds
    Bicubic rendering... 1.204 seconds
Score: 2827

Running 'Database' benchmark...
    Create database... 0.624 seconds
    Populate database... 2.137 seconds
    Save database... 0.274 seconds
    Reload database... 0.078 seconds
    Search database... 1.861 seconds
    Sort database... 1.650 seconds
    Group database... 1.055 seconds
Score: 2800

JRMark (version 18.0.136): 2350
So I don't think it is just a weaksauce VM machine.  This is, after all, running on a dual CPU, quad-core Xeon (2.26 GHz) Mac Pro with 16GB of RAM (3.5GB allocated to the VM).

I'm ONLY seeing this kind of behavior on this particular machine, but I use it as part of my regular workflow at the office, and it is annoying (and seems to be getting worse).  Is there any tips I can use to diagnose this problem?

I never see CPU activity spike while it is "frozen".  It absolutely feels like it is either a stuck process that is returning some kind of error, or a disk-latency issue.  But, other than having the files themselves on a faster disk (which isn't possible), I can't think of what I can do to improve the situation, and I've tried a bunch of things.

I'm happy to collect additional logs showing it if it helps.  It is quite easy to reproduce.  I happens constantly, and it seems to be getting "worse".

[bivan]

These lines look interesting:

0028203: 1936: General: CFileSystemManager::AnalyzeDrive: Error using connection '\\quicktime.jax.org\movies': 67
0028203: 1936: General: CFileSystemManager::AnalyzeDrive: Drive does not exist (drive: \\quicktime.jax.org\movies, filename: \\quicktime.jax.org\movies\samples\sample.mp3, type: 0, attributes: 0)

0089438: 5360: General: CFileSystemManager::AnalyzeDrive: Error using connection '\\wallvideo.jax.org\library_data': 67
0089438: 5360: General: CFileSystemManager::AnalyzeDrive: Drive does not exist (drive: \\wallvideo.jax.org\library_data, filename: \\wallvideo.jax.org\library_data\test\Sample.wtv, type: 0, attributes: 0)

0110469: 5360: General: CFileSystemManager::AnalyzeDrive: Error using connection '\\wallvideo.jax.org\Users': 67
0110469: 5360: General: CFileSystemManager::AnalyzeDrive: Drive does not exist (drive: \\wallvideo.jax.org\Users, filename: \\wallvideo.jax.org\Users\Public\Pictures\Sample Pictures\Chrysanthemum.jpg, type: 0, attributes: 0)

[InflatableMouse]

Is it running on Mac OSX in Fusion or Parallels, or do you have Windows running on your Mac Pro?

What's the guest OS?

The only thing I've ever encountered that comes somewhat close is clipboard sharing. That was on Windows guest and host on Vmware Workstation and I had to disable clipboard sharing. With the guest inactive, putting certain content on the clipboard would make Vmware unresponsive, not just the guest.

I guess it doesn't hurt to try to disable integration features although I must admit I'd be surprised if that was it.

[csimon]

Yep, I see this happening sometimes.  But it's only when the NAS on which I store my media files is off. So it's the usual Windows thing of hanging when a drive or network share is inaccessible.

[Cmagic]

Hi,

I'm also seeing the same kind of behaviour on my home iMac.
I run MC18 in a Parallels 7 VM and MC freezes from time to time.
I am generally able to fix this by changing the VM network mode. I'm usually in a bridged network configuration (because that's the only mode where media server works for me)
When MC freezes, I switch temporarly to shared network and then switch back to bridged as a result MC resumes to normal operations. Well, most of the time.
So it looks like it is some kind of network access problem.

cheers,

C.

[robertj388]

I am running MC 18 under Parallels using a MacPro.  Desktop speakers work just fine, but if I try and send sound using DNLA to my receiver, I get a message that MC is "waiting" to send sound but it never does.  The receiver shows a "server error".  The same thing happens under windows 7 and 8.  Previously, it worked fine with windows 7, but now both instances cause the "waiting" message to ococur.  Again, switching back to the player for the computer desktop, everything works just fine. Thanks so much for helping.

[Fred1]

I have this behavior, too, on a native windows machine when doing tag updates, analyzing etc. on a big library.
The GUI freezes and i can't do anything but wait.