INTERACT FORUM

Please login or register.

Login with username, password and session length
Advanced search  
Pages: [1]   Go Down

Author Topic: Workflow for import and tagging new media files  (Read 1222 times)

HTPC-Freak

  • Recent member
  • *
  • Posts: 39
Workflow for import and tagging new media files
« on: December 08, 2022, 12:25:06 pm »

Hello,
I hope to use JRiver from now on in a more structured way as before - for that I have read many, many posts in the wiki and in the Forum to learn before doing things wrong...

The Goal:
Develop a practical, safe and efficient workflow to import many various media files and tag them in JRiver

Computer setup:
1.) Headless play out client PC (Windows 10, without keyboard and monitor). This PC is connected with audio system and monitor for presentation only. it is used with Android tablet and Gizmo remote control app

2.) Office workstation with two monitors running JRiver on Windows 10, This PC is used for tagging and maintain the JRiver database only, not for play out

3.) the media content is stored on a NAS (Unraid Server, used for many other services as well) This Server is only used as a NAS for JRiver

I am still running latest MC29 on Office workstation and Headless Play out Client PC

Here is my first question:
- can I tag the Media files, without the NAS ? So are the tags fields in the database or are they part of the content itself? So if the content is not accessible, because the NAS is switched off, can I still modify the JRiver Database (Main Database on the client play out machine) and add some tags ?
- For security reasons, I would like to limit write permissions to the NAS as much as possible. Is it necessary to grant write permissions for the client play out PC, which holds the Database to the NAS when tagging ?

Is it necessary to manually save the database after tagging, or is saving the database an automated process ? Somewhere I read, that the database is automatically saved, when power off JRiver software, however it seems to me, that this is not working properly. The saved Database zip file although showing timestamp of late evening, did not contain tagging work from the many hours work of that day before... frustrating loss of work. Did I misunderstand something with automatic database backups
So manual database backup is mandatory after tagging ?

After tagging is complete it is time to empty the folder "recently imported" and sort the media files in Audi/Video/.... categories on the left, I think
But before, I want to know the answers to my questions above.

Is there a step by step workflow explained somewhere ? I haven't found it, so I thought it is useful to document the development of my workflow here, considering the setup of dedicated Play out / Maintenance and storage infrastructure in my media network

regards and thank you for continuous support
Logged

zybex

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 2616
Re: Workflow for import and tagging new media files
« Reply #1 on: December 08, 2022, 02:46:39 pm »

Hello,
Hello!

Quote
Here is my first question:
- can I tag the Media files, without the NAS ? So are the tags fields in the database or are they part of the content itself? So if the content is not accessible, because the NAS is switched off, can I still modify the JRiver Database (Main Database on the client play out machine) and add some tags ?
I just wrote this yesterday, which is useful background:
https://yabb.jriver.com/interact/index.php/topic,134623.msg932500.html#msg932500
MC really likes the media files to be accessible. If you disable tagging to the files and XML, you can do tagging on the DB only and later on do an "Update tags from library" to push the tags to the files, if you wish. However, you'll still see MC freeze ocasionally for 30 seconds when it's trying to access the NAS for some reason and it's not available. This is annoying but can be mitigated with some config changes.

Quote
- For security reasons, I would like to limit write permissions to the NAS as much as possible. Is it necessary to grant write permissions for the client play out PC, which holds the Database to the NAS when tagging ?
Assuming your "Workstation" is configured as the "Media Network" Server, your client machine can be connected in Read Only mode to the Server. It won't be able to change [almost] anything, just play. You can also have the client connected to the server in full mode for more flexibility and limit it to read only in the NAS. MC will work fine. Some things like [Number Plays] tag won't be updated in Read Only mode.

Quote
Is it necessary to manually save the database after tagging, or is saving the database an automated process ? Somewhere I read, that the database is automatically saved, when power off JRiver software, however it seems to me, that this is not working properly. The saved Database zip file although showing timestamp of late evening, did not contain tagging work from the many hours work of that day before... frustrating loss of work. Did I misunderstand something with automatic database backups
So manual database backup is mandatory after tagging ?
You seem to be mixing "saving" and "backup". Saves are automatic - as soon as you change a tag, it's written to the DB and to the media file (if enabled). Backups as copies of the DB performed at regular intervals so that if something goes wrong you don't lose too much work. You can schedule more frequent backups in MC if that makes you happy.
Also, note that the MC Server (on your Workstation) is actually still running when you close it in order to keep acting as a server. There's a tray icon for it, and you can see it in Task Manager too. So if you schedule a task to do a backup "on application exit", it won't actually do what you expect it to, because it's not actually exiting.

Quote
After tagging is complete it is time to empty the folder "recently imported" and sort the media files in Audi/Video/.... categories on the left, I think
But before, I want to know the answers to my questions above.
There's no need to empty it.

Quote
Is there a step by step workflow explained somewhere ? I haven't found it, so I thought it is useful to document the development of my workflow here, considering the setup of dedicated Play out / Maintenance and storage infrastructure in my media network
No. Though this is asked so often that it may depend on your search skills on this forum and wiki.

Quote
regards and thank you for continuous support
You're welcome.
https://www.youtube.com/watch?v=eBGIQ7ZuuiU
Logged

lepa

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 2033
Re: Workflow for import and tagging new media files
« Reply #2 on: December 08, 2022, 03:05:46 pm »

However, you'll still see MC freeze ocasionally for 30 seconds when it's trying to access the NAS for some reason and it's not available. This is annoying but can be mitigated with some config changes.
A bit off topic to OP but I got rid of MC accessing unavailable NAS files in Standard View by not using drive letters (Windows) but just network paths AND moving all the cover files to always on SSD outside the NAS.

Unfortunately you'll need 3rd party SW to achieve the latter as MC always saves video cover files next to media file
Logged

eve

  • Citizen of the Universe
  • *****
  • Posts: 689
Re: Workflow for import and tagging new media files
« Reply #3 on: December 13, 2022, 02:56:20 pm »

A bit off topic to OP but I got rid of MC accessing unavailable NAS files in Standard View by not using drive letters (Windows) but just network paths AND moving all the cover files to always on SSD outside the NAS.

Unfortunately you'll need 3rd party SW to achieve the latter as MC always saves video cover files next to media file

I actually handle this sort of hybrid storage approach regularly for my media library.

Use MergerFS on your NAS.
Media is on slow, bulk storage
Metadata is on a fast SSD.

Metadata SSD's content lays over the bulk media storage which is 'read only'. When JRiver or other software changes metadata, generates or scrapes images etc, it'll write to the SSD, not the HDDs.




Logged

lepa

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 2033
Re: Workflow for import and tagging new media files
« Reply #4 on: December 13, 2022, 04:32:43 pm »

You also just described my configuration too ;D

Although in addition to cover art I also have my music media files on local SSD as they are in daily use. All the other media files are on hibernating snapraid+mergerfs server which is scrubbed and synced when server is in use. MC also only runs in the client so I don't need to use MC's bit restricting client-server setup.

I was having really really bad MC not responding lock ups all the time before I utilized your tip not to use drive letters even though all the cover art was already moved away from server so there shouldn't have been any reason to access files when just changing view etc. Anyway, now I'm a happy camper even though 3rd party programs are needed to maintain the peace
Logged

eve

  • Citizen of the Universe
  • *****
  • Posts: 689
Re: Workflow for import and tagging new media files
« Reply #5 on: December 14, 2022, 02:39:16 am »

You also just described my configuration too ;D

Although in addition to cover art I also have my music media files on local SSD as they are in daily use. All the other media files are on hibernating snapraid+mergerfs server which is scrubbed and synced when server is in use. MC also only runs in the client so I don't need to use MC's bit restricting client-server setup.

I was having really really bad MC not responding lock ups all the time before I utilized your tip not to use drive letters even though all the cover art was already moved away from server so there shouldn't have been any reason to access files when just changing view etc. Anyway, now I'm a happy camper even though 3rd party programs are needed to maintain the peace

Glad that tip cleared up your performance a bit :) Even with sane mergerfs configs and all your metadata on SSDs, certain things will 'ask' your regular storage to list it's directory content (which can also include things like the 'file size' of each file in a directory) or access it in other ways. Now in *most* situations, this isn't going to be a big problem. You're dealing with extremely expansive datasets? Yeah it starts to actually make sense to think about how to approach how we do this. For example, some of my database backend stuff has no access to media itself, it can only see the metadata stored on an SSD. In testing this proved more performant than the same operations happening on the hybrid mount.
Also windows in general sucks at trying to read way more than it needs to from directories (for example, you may have thumbnailing off but that doesn't mean windows may not try!)
Logged
Pages: [1]   Go Up