Configure player

Close

WWDC Index does not host video files

If you have access to video files, you can configure a URL pattern to be used in a video player.

URL pattern

preview

Use any of these variables in your URL pattern, the pattern is stored in your browsers' local storage.

$id
ID of session: wwdc2008-543
$eventId
ID of event: wwdc2008
$eventContentId
ID of session without event part: 543
$eventShortId
Shortened ID of event: wwdc08
$year
Year of session: 2008
$extension
Extension of original filename: m4v
$filenameAlmostEvery
Filename from "(Almost) Every..." gist: [2008] [Session 543] Deploying F...

WWDC08 • Session 543

Deploying Final Cut Server

Integration • 1:15:43

Final Cut Server is Apple's powerful new media asset management and workflow automation tool. Learn the best practices for deploying Final Cut Server. Discover how to configure the event-based response model to build complex workflows that can track job status, and monitor media changes, media delivery and archiving.

Speaker: Matthew McManus

Unlisted on Apple Developer site

Downloads from Apple

SD Video (400.7 MB)

Transcript

This transcript was generated using Whisper, it has known transcription errors. We are working on an improved version.

Good afternoon, everyone. My name is Matt McManuus, and I'm from the Final Cut Server engineering team. So today we're going to talk about deploying Final Cut Server. So Final Cut Server is a new technology from Apple. Just to give you an idea, how many people actually have Final Cut Server now? So about half of you, so you probably have a few questions about how you might actually deploy it and use it. And for the rest of you, you probably have some questions about exactly what it is and what it does.

So today, I'm going to go over two main objectives. The first one is to teach you about the mechanisms that are involved in deploying Final Cut Server. And secondly, I'll show you some steps involved in an actual deployment. So to do these, I'm going to go through the following areas.

I'll give you a brief introduction to the product, and then I'll go over the key components of Final Cut Server that are used in a deployment. Then we'll talk about the seven steps of deploying Final Cut Server. I'll talk about post deployment, so the care and feeding of Final Cut Server. Then we'll talk about some other considerations. And finally, we'll have some Q&A.

So Final Cut Server was designed to address two key areas. The first one is media asset management. And the second is workflow automation. So we'll talk about the media asset management first. So the problem there is just the huge amount of content that you find in a typical installation these days. Maybe it's on an XSAN or on a FireWire drive on a shelf somewhere. But more and more, there's all this file-based media to deal with, and even other media as well. So Final Cut Server enables media asset cataloging.

And that comprises of a few key areas. First of all, Final Cut Server is able to catalog all file types. So any file type at all, not just media files. It can also catalog Word documents, PDFs, graphics, anything. And when it catalogs them, it creates what we call an asset out of it. And we'll see that in a bit more detail later, as well as how you get stuff in and out of Final Cut Server.

We also have customizable metadata for all of our assets. So you can actually choose what metadata you have for a given asset, and you can add additional fields. And finally, there's also the ability to read metadata from both QuickTime files and XMP metadata from other file formats. For certain file types, we're able to automatically generate small clip proxies and thumbnails.

So in the case of a QuickTime file, will use Compressor to make a H.264 smaller version of it, so you can actually use those proxies to browse quickly and easily. And we have the ability to synchronize automatically with the file system. So you can add devices to Final Cut Server, which we'll see. And then as files are added to those devices, Final Cut Server can automatically catalog them and add assets for the new files.

Now once you've got everything catalogued, you need to be able to find it. So in Final Cut Server we've got the ability to have some powerful search tools that allow you to search on either a given metadata field, or where an asset is located, or a number of other different parameters.

There's also the ability to have some access controls and access restrictions, so you can limit what a given user can see and do with the system. There's also the ability to then integrate with archival systems. So you can move assets from, say, an online XN environment to a tape-based archive.

Now, workflow automation is another key area. And for this, we have what we call a watch and respond system. And the idea is that we have these actions or responses, for example, a copy or a move to a device. And then we can apply those actions to various events. For example, if I change a piece of metadata, it might trigger actions which will copy, convert to a H.264, send to a web server, and send an email.

There's a simple way to configure some of the workflow, and we'll see how we can do that from the Assistant Preference pane. And as required, you can build out some much more complex workflows in our advanced administration area. Now another key area is collaboration. So Final Cut Server enables users to collaborate.

There's the ability to have checking in and checking out of assets, including Final Cut Pro projects.

[Transcript missing]

which then enables me to revert back to a previous version as required. And you can also set up automated notifications, so you can be notified when, say, a project is checked in. And then you can use these to build out a review and approve process.

Final Cut Server is deployed typically in an XServe environment. The requirement is Leopard. So Leopard Server or Leopard Client. So for sort of mid-size to larger size installations, there might be an XServe with a compressor farm attached for the transcoding. and perhaps a XAN backend or attached RAID storage.

But the design of it is that it can also run in small work groups. So if you've just got a few users, or if you're a user who has, say, a Mac Pro and a laptop, and you want to catalog and share files and use some of the features in Final Cut Server, you can install it just on a Leopard client onto a Mac Pro.

It's a client-server-based architecture. So you've got the server, which installs on Leopard, but the client is Java-based. And that can run on either Tiger or Leopard, but it can also run on Windows XP or Vista. So it enables your Windows users to actually access the assets and participate in perhaps the review and approve process.

I'll talk a bit more about the architecture. Once you've got Final Cut Server installed, You then connect it to devices. So the devices might be XSAN or a network file server, or perhaps some local storage. And once you've connected to these devices, the Java client accesses Final Cut Server. And then can upload assets or move assets from device to device, or perhaps download assets to their local machine, or check out projects to work on them locally.

and this is all done over the Ethernet network. Now, if you've got remote users, you can give them access to your assets as well. And the recommended approach to do this is using a VPN. We do not recommend putting Final Cut Server onto the open internet. Your assets are a bit too valuable to be able to risk that type of access.

So to talk a bit more about what's actually part of Final Cut Server, there's three main processes that are installed. We've got this when you're on the actual server. So we've got the store D, or what we call the store server. That's the brains of Final Cut Server. So that coordinates all of the client interaction.

and the other things that it's actually doing. It's got a PostgreSQL backend database, so all of the metadata for the assets and configuration and other things are stored in that database. And there's also the LDS, or the device server. And that takes care of the heavy lifting, so the talking to devices and to the transcoding engines.

So to give you a bit more detail, the client, the Java client when it's run up will talk to the store server or store D, and all of the communication for the searching or the changing of metadata is done through that, talking to that process. And then the delivery of assets or the thumbnails comes from the LDS. Now Compressor is, which does all of the transcoding in Final Cut Server, will talk to the LDS. And the LDS also talks to the various devices. So any exchange with devices or moving of content is done by that process.

So, there's two ways to administer Final Cut Server. The first way is through the system preference pane. And that's available on the server on which you installed the Final Cut Server server.

[Transcript missing]

You can access an administration area. And this allows you to do some much more advanced administration.

It's very flexible, but that flexibility does come with some complexity. So what you might find is as you deploy, you might do a subset of deployment actions in the system preference pane. For example, add devices or add some simple automations. And then you might tweak or add some more advanced Workflows or change metadata in the Java administration. So now I'll talk about some of the key components of Final Cut Server.

So firstly, we've got devices, and then we've got the assets, productions, metadata,

[Transcript missing]

So the devices, as we've seen, can be added to Final Cut Server. and the devices can be of a few different types. There's a local file system, so you can add any directory on the local machine as a device.

You can also add a XSAN as a device, as well as the network file systems. So there's support for AFP, SMB, NFS, and FTP. And finally, there's a special device type in Final Cut Server, and that's the content base. Now the Content Base is a managed device which allows files to be stored in a non-hierarchical way. and it also allows for files to be stored without worrying about file name collisions. Essentially, it's a bucket for files.

So examples of how it's used is in Final Cut Server there's a device which is set up to store all of the proxies which are generated for assets. So that proxy's device is a content base. In addition, the versions. So all of the versions of an asset are stored in another content base.

An example of how you might actually use a content base in a production is for a graphics library. So you might have a graphics library which might have, say, 50,000, 100,000 images, and you can just add them to a device which is a content base without having to worry about the file structure or... Worrying about if, you know, this file name is going to overwrite this one. It's all taken care of by Final Cut Server.

So there's also a special idea that we've got in Final Cut Server with Advices, and that's edit in place. So when you've got a device which is not edit in place, and the user tries to use a file, the file will be cached on a local machine. What edit in place allows is if a user wants to use a file, for example, they're using a QuickTime file in a Final Cut Pro project, The edit in place allows a link directly to the asset on a mounted file system. So the key example here is XAN. So when you add an XAN device, edit in place is set up automatically.

So when the file is used, it's not cached, there's nothing which goes across the network. A direct link is provided to where the file is on the XAN. Now, it is possible to add edit-in-place paths

[Transcript missing]

You can actually set that up for edit in place, so when users use the files, they're not downloading locally. They're pointing directly to where they are on the AFP share.

and archive devices. So you can actually set up any of the device types as an archive device. For example, SMB, AFP, all of these, when you add a device, there's the checkbox you want to enable as an archive. So once you've actually got an archive device, it becomes a target for moving a file, is the founder of XAN, a software development company that has been working on the XAN system for over a decade. So now let's talk a bit about assets. So once you've got these devices set up, You can actually start to create assets. So an asset is created for every file that is catalogued.

An asset has a few very key things about it. So it's got all the metadata. So we can see here some various metadata. So I've got title, category, description, plus any additional metadata fields that you want to add to the asset. In addition, the asset has resources. So it has links to files.

The most important link for a given asset is what we call the primary representation. And this was the file that caused the asset to be created in the first place. So it also has other resources, which are the proxy representations. So for example, here we've got a clip which has a thumbnail, a poster frame, and a clip proxy. So to read it, so the asset, every asset has metadata and it has these resources.

An asset, you can actually check it in and check it out using the client. And you can also archive an asset. There's a special class of asset which are created for Final Cut Pro projects. So when you upload a Final Cut Pro project to the system, it creates this Final Cut Pro project asset. And the key difference with these assets is they have what we call elements. And the element corresponds to the contents of the bin of the Final Cut Pro project.

and each element has a link to a master asset. So when you upload a Final Cut Pro project, any assets which aren't already managed by Final Cut Server will be created as new assets. And the elements will be added to the Final Cut Pro project asset. So here we can see I've got an element which has a link back to a master asset. And it has, tells me what project it's being used in.

So, before I move off assets, I'll just be clear about how they can actually be generated. So, you can do it in a few different ways. So, for a given device, you can set up what we call a scan, and a scan will just catalog or find any content on that device or on a subfolder on that device. and you can set up these scans on some kind of schedule. So perhaps every five minutes, every day, whatever you desire. So every new file that these scans find, it creates the asset.

You can also create assets when you actually upload a file using the Java client. You can create assets when you put A special folder, which we call a watch folder, and that can generate assets. And we'll see a bit of that later on. And finally, assets are generated through this mechanism, through uploading a Final Cut Pro project. Final Cut Server analyzes that project. It sees what media files are being used in the project. and it uploads them to the device if they're not already on a managed device, and it creates assets for them.

So that's assets. Now we'll talk a bit about productions. and productions in Final Cut Server is just a container for assets. So you can make a new production, and you can just add assets to it, either manually, drag and drop, or when you upload, you can choose, I want to add these assets into a given production. So these just allow you to collect assets for a given task. So for perhaps a show or some other little project, a production can be made to collect all of that. And just like assets, they have metadata associated with them.

So let's actually talk about that metadata. Metadata is very key to Final Cut Server. So we've seen that the metadata can be used for either assets or productions, but it also drives a lot of actions inside of Final Cut Server. One might say it's a metadata-driven piece of software. The key about metadata is there are a few different levels when you actually construct metadata for assets or productions.

At the top level are metadata sets. So a given asset or production belongs to a metadata set. So when I upload a QuickTime file, I actually choose what set I'm gonna use for that asset. or when I do a scan, I'm actually deciding for all the files which are cataloged from the scan, what metadata set they're going to belong to. So sets are at the top level. Now sets can contain groups. And you can have multiple groups for a given set.

And then to the groups, you can actually add the fields. So the description, keywords, job number, So you've got the sets, the groups, and the fields. And then for the actual, for the fields, you can have lookups. And so a lookup allows you to have a pull-down, and then allows you to do a few things. So you can actually restrict the choices available for a given metadata field. It can simplify the searching when users want to find assets. and it can also ensure that automations are actually triggered correctly. So you might have a given automation which is triggered off a given lookup.

Now, metadata is also used to do a few other special things. So there's some special metadata groups that allow you to do things like

[Transcript missing]

And to administer the metadata, you've got to use the client administration. So if you want to add fields, so forth and so on, it's the client administration.

And another key component is the ability to then map from metadata in, as we saw, QuickTime or XMP, into internal fields and back out. If that's an area of interest to you, then I can highly recommend a talk we've got tomorrow, which will be discussing that in some great depth.

So now let's talk about access restrictions. So with access restrictions, you can limit what a given user can actually do with the system. So the way these work is that you define permission sets in Final Cut Server. and the permission set can restrict access in the following ways. The first way is by restricting on asset metadata.

So I can say that this given permission set can only see things which are either, say, approved, ready for review, or completed. So when the status equals any of those three things, then the user who is tied to that permission set can only see those assets. And I'll show you some of that later. In addition, you can restrict on production metadata. is the founder of Final Cut Server.

He's been working with us for over 20 years. He's You can also restrict on devices themselves. So for a given device, you can say that this permission set cannot delete from this device, cannot even see assets on this device, cannot create assets on this device. And finally, there's the ability to restrict based upon traits. And this allows you to do some things like turn off given metadata sets for a given permission set.

Once you've actually got some permission sets built, and Final Cut Server has about five or six out of the box, then you can actually tie the permission sets to Mac users and groups. So on Leopard Server, this would tie directly into Open Directory, and on Leopard Client, You can do the same sort of thing using the users and groups in the system preference pane.

So a user who's been put into one of these groups, will then get the permission set that's been mapped to that group. And again, I'll show you some of that a bit later on. The next key area I want to talk about is transcoding. So Final Cut Server uses Compressor for all of the QuickTime and audio conversions. The great thing about this is you can actually set up a compressor cluster to increase the speed of the transcoding.

And Final Cut Server handles all of the still image conversion locally by itself. - I'm using some system processes, but the idea is that if you want to farm off all of the QuickTime and audio conversions that can be put off to a compressor cluster, and the Final Cut Server machine will then take care of just the image transcoding.

You can make custom compressor settings in Compressor, and these can be added to Final Cut Server. And you can also restrict what compressor settings are actually used by Final Cut Server itself. So you can say that in a given deployment, we're only actually going to use four or five different settings. So you can actually turn off the other settings. or perhaps as a play out server which needs a particular kind of QuickTime or some other format. So you can make sure that any copies to that device only have that particular setting available.

and he can also extend Compressor with the use of plugins. So if you wanted to go out to different formats, for example, Flash, MXF, GXF, you could use the Episode Pro plugin from Telestream. Or if you wanted to write out, say, a DPX sequence, you can use a Glue Tools plugin. Finally, you can use QuickTime components to extend the ability to read certain formats. So one example here is with Windows Media. So you can actually use the flip for Mac. QuickTime component to enable reading of Windows media files.

Finally, I want to talk about automations. Automations are tremendously powerful and enable you to build out the workflows, but there's quite a bit to them. The idea is that you can create responses to perform actions, So copy and delete, or email, and we'll see some examples of those.

And then the responses are triggered from either a watcher, which is just a folder on a device which Final Cut Server is looking at, or a schedule, which is going to do something on either a periodic or an absolute point in time. And finally, my favorite, subscriptions. And we'll go into a lot more detail on subscriptions. So, you can create some simple automations, as we've seen using the system preference pane, and as we've discussed, go into far more detail using the client administration.

So let's talk a bit about responses. There are 14 different responses available. For example, copy. So this response enables you to copy to a given device and optionally transcode. So I could copy to a web server and transcode to H264. Or copy to my Playout server and transcode to ProRes. There's the ability to send an email. So you can send an email to a given user. and the powerful thing about this response is it has metadata substitution. So you can actually substitute in metadata from the asset to build the email.

There's also a scan response, which I've discussed, and this can be set up on a schedule to catalog content on a given device. and an archive response, so this can be triggered to move an asset to an archive. One example of this is you can actually have a, set up a system to be able to, say, archive everything older or everything which has not been accessed for, say, over 30 days on the XAN to archive.

So you can build out these sort of workflows. And there's also the delete response, which you can attach, which will delete a file. A set metadata response, so you can actually set metadata on a given asset. and my favorite, the execute command or script response. So this enables you to execute a executable on the local server. and again we'll be going into much more detail on this execute command or script in our integration talk tomorrow at 10:30.

And there are about seven or eight other responses that you can call upon. So once you've actually built out these responses, they can be triggered in a number of ways. So we've seen the watches. So it's a Any new or modified file on a folder or subfolder of a device? can then do these various actions. So there might be a watch folder set up. to then convert to a given format and send to a device. or to go into a content base, send an email to the reviewer to review it.

There's also the subscription. So this is where we subscribe to events on assets or productions. and we can filter on metadata. Now let me go into a bit more detail here. So we've got the idea of events. Now, an event is something like asset creation, or asset modification or asset deletion. So these are events in Final Cut Server. So when you set up a subscription, it's actually looking for these events. So it's looking when assets are modified.

And then you can further pare down exactly what it's going to trigger based upon metadata. So an example which I'll show you later is to be able to... sent out an email based upon when an asset is set to say ready for review. So there's a subscription to when an asset is modified, and it's looking for when my status field has changed to ready for review. And when all of that is matched, then it actually does these responses which have been attached to it. And finally, there's the schedule, which is, as I said, can run on an absolute or periodic time.

So that's the key components of Final Cut Server. Now I'm going to talk a little about deploying. And I've got here seven steps of deploying Final Cut Server. These are the steps which I've built up based upon my experience with both Final Cut Server and its predecessors. and they're certainly not to hard and fast rules. But what they allow is you to build on the previous steps to actually get to a successful deployment. So the first step is perhaps the most critical one, and that's actually planning.

Because the actual, you'll sort of see as we go through this, that there's all these other things that you need to configure, and if you actually know what you're trying to do before you install the software, you're much better off. Now the next step is just the installation and initial configuration. So this is installing the software onto an XServe or a Mac Pro. and doing some initial configurations. So, setting up a compressor cluster and connecting up to it.

or setting up what your outgoing SMTP server is. Just these initial configuration steps. Then it's a matter of configuring the metadata and transcode settings. So it's a matter of working out what metadata you want to track, you know, what metadata is used in the facility at the moment, and what you want to have in the assets.

And then what Transcode settings do you want to use? So what are the house settings? What are the settings that are required for my different devices? All of that. Then it's adding devices and setting up scans for those devices. And the reason that I do it this way is that when I add devices and setting up scans, I'm actually going to be using the metadata fields and the transcode settings that I created in the previous step.

So once the devices are set up, and the scans have started to happen, you can then add, create the permission sets. and then configure your open directory users to then tie users to those permission sets. Then the next step is to start configuring the workflows. So, to set up the watch folders, or to set up the various responses and triggers, And finally, once all that's done, configure the backups and various notifications that you want to get as an admin.

This is not to say that once you've actually gone through this deployment that you can't change anything later. There's plenty of post-configuration that can be done, and tuning which is done to actually get to the kind of system that you're after. But these are just some steps to enable you to go through the initial deployment.

So let's talk about planning first. So the key point here is there's certain questions that You need to be able to answer before you can move on. So that's things like, you know, what are the devices that I'm going to connect to? What's the various underlying infrastructure? Then what transcode formats am I going to use? Or what various formats are going to be used in the facility? What are the users and groups? So you might have, say, some producers and some editors and some managers. So it's a matter of thinking about what sort of access do you want to give to these people? So you want to say the editors will want to have access to perhaps all of the raw material and other bits of material.

And then the producers may only see things which have been completed or ready for review, or perhaps an education environment. You can have different groups for each class. and add the users to a given class, and they can only see the content which has been set up for that class.

Then it's a matter of asking, you know, what metadata do I want for my assets, and what mappings do I want to have from external formats? So if you're using, for example, a P2 camera or some other file-based camera which supports metadata, do I want to actually see those fields in my assets? Which ones do I want to get? How do I want that to work? What are the workflows? How does a job get from inception to completion? or how do I want things to be submitted, or what's my review and approve cycle? Finally, you need to work out what the backup and recovery policy will be.

And I'll talk a bit more about backup later. But that's, you know, how are you going to get the system back up and running after a fire? I've got an example here, which is Acme Productions. Just a typical production house. They've got some Final Cut Pro editors, they've got some producers using either MacBooks or Windows laptops, and they need to output for a Playout server and a website.

So if I was actually going to plan this deployment, one of the first things I'll do is actually make a system diagram, which will sort of outline the various devices. And the great thing here is I can very easily see, oh, okay, I need to add my web server as a device, my Playout server. It needs these formats. And, oh, my XAN. And, oh, there's my compressor. All that sort of stuff. These allow you to sort of quickly identify, okay, here are my devices, here's how it's going to set up.

Then I'll build out some sort of workflow diagram, so sketch out what the workflow is. Often I'll use paper and pen, but I might sort of have, so here's how a job gets from start to finish. And here are the various forks or various things. And so when I actually build out these workflow diagrams, I can quickly identify, oh, I need an email response here, and it's got to be triggered by this piece of metadata.

Or perhaps I need a copy response here to then go out and convert to this format. So, it's by diagramming the system and then talking to the users to then work out the workflow, that you can then start to build out a list of the various things that you want to configure in Final Cut Server.

So now I'm actually going to show you some of the next steps, and we can see how they are on a real Final Cut Server client server thing. So here is Final Cut Server. I have it installed already. The key note about the installation is when you install, there are a few key things you need to decide.

There's the idea of different profiles. There's like four or five different profiles. Film, production, education. What these do is they set up different metadata sets and groups. So we've seen what they are. They just allow for subtly different profiles. is the founder of Final Cut Server. He's a great guy, he's a great guy.

He's a great guy. And the other key question in the installation process is where are my proxies going to go, and where is my default media going to go? Now, I'll talk about this a bit later on, but the key answer there is it's very important. It needs to go on some rated storage. So either system drives on an XSERV which are rated, or on a direct attached promise RAID unit or some other RAID unit, or on some sort of XSAN.

So here we have Final Cut Server. So this is the client for Final Cut Server. and we can see here that I've got my various assets. So I've actually got A few devices already connected to this server and I've actually started to scan those devices. And so if I have a look at one of these, I've got my asset with my various metadata. and I can use my search tools to be able to find things. So if I wanted to find all of my, say, pipeline masters material, I can use just my search tools to find all my pipeline material.

And once I've actually built out a search, I can also, here I've got an advanced search area. So I can choose particular fields that I want to actually search upon. So I can say, you know, I want to see all my pipeline material which is stored on my XN.

Or I'd like to see any assets which have been created in the last two hours. So once you actually start to build these searches, you can actually save them. So I can save my pipeline search as a smart search. So I've got a button down the bottom right-hand corner. Save a smart search. And it can say pipeline.

And so now I've got my save search. Or I can see some other save searches, my raw files and my assets, so forth and so on. Now what I want to do is my step three, which is configure metadata and transcode settings. So right now, I have some metadata for my assets.

I've got my different metadata groups. Here are the groups. I've got asset, I've got a P2 info group, where I'm mapping in is the founder and CEO of Final Cut Server. He's been working on a number of new projects, including the first one, which is called Final Cut Server.

He's been working on a number of new projects, including the first one, which is called Final Cut Server. He's been working on a number of new projects, including the first one, which is called Final Cut Server. So let's say I wanted to add some fields and groups. I can quite easily do that by going to my Java administration over here.

I get a little warning saying, "Do I really want to do this?" Yes, I do. I think I know what I'm doing. And so I've got the option here to add a "My fields and my groups and do various mappings as well as create lookups." So let's say I wanted to create a field in the group for my Acme productions. So I can quite, let's say I wanted to have a job number and also have a client lookup. So I've got different clients, so I want to give the clients different access. So to do that, I can... create a metadata field and call it, say, job number.

I've got different options for my data type, so it might be a Unicode or perhaps in this case it's going to be an integer. Give it different categories. So I've got my job number field, and now I can add another field which might be a client. I'll give it category of none. Fantastic. And then for once I actually have these set up, I can add different lookups. And it can make a lookup for client. And it might have, you know, XYZ. XYZ.

Another one called ABC. ABC. So I've got two simple clients there. Save Changes. And now I can create my group for Acme. Make a metadata group, Acme. I can then choose my fields. So I've got my job number field, and my client field. And to the client, I can actually assign it the lookup which I had just made. So I can say, field lookup, client.

I can add some actions. And then I can say that it's going to be in my media metadata group. So if I do that, and now have a look at one of my media assets, I'll now see I've got my Acme. Metadata Group. This asset has metadata set of media, And if I look at my Acme group, I'll see I've now got my job number.

[Transcript missing]

So that is the creating of the metadata. Then it can create transcode settings, if I so chose, in Compressor and add them in. We're running a bit short on time, so I'm going to keep pushing through. So then the next step is the adding of devices. And as I said, the two different ways of the administration. So if I go into-- My system preferences. I've got my Final Cut Server administration.

And so I've got a few different options here. So I can assign my group permissions. So here I've got my permission sets which I've created in Final Cut Server. And these are the local groups. And I can also add devices here. So let's say I wanted to add a -- this content folder on my desktop as a device. So I can add Hit Plus, and here we can see the different device types.

So I've got local network or XN. So this is going to be a local device. I can call it content. And I can just drag in my local directory. So now I've got it added as a device. And in the next step, I can actually say, is it an archive device? No, not in this case, but this is where I would set that up.

Here I can set up the scan settings. Now, when I set up a scan in the preference pane, I'm scanning the entire device. If I set up scans using the client administration, I have the ability to scan different subfolders of a device. Now the reason why that's important is when I set up this scan here, I do an ad-only scan every minute. I'm setting up a metadata set. So these are the various sets that I have available. Audio, document, graphic, media project. And I can make additional sets, but all of the assets on this device are going to be cataloged and added to my media metadata set.

So, if I wanted to have a bit more fine-grained control, What I'm able to do is to, in the... Client administration to be able to specify a subfolder on a device. And for that subfolder, I can assign a different metadata set. So for example, on an XN, I might have graphics, and I might have clips, and I might have audio.

And I want different metadata sets for each of those classes of asset. So I'd set up different scans pointing at my subfolders of my XN. In addition, when I set up scans in the Java administration, I can do additional things like set up filters. So I can say, only scan things which match something ending in PSD or MOV.

So if you want to scan items on a given device and you only want to grab the QuickTime files or the other "If you want to use content, you know, particular content, you can restrict the scans in that way." So the other, so once we've actually added a device, so I've actually set up my scan here. And in about a minute, I'll actually see the scan progressing in my job page. So the job page shows everything that the server is doing.

So whilst I'm waiting for that, I can actually start to think about my users. So if I want to say I've got my client, I only want to see their stuff. So stuff which is XYZ, which is marked as ready for review. So that's all I want my clients to see.

So I'm just going to mark two or three assets like that now. Ready for review and client is XYZ. So what I can then do is build a permission set which then restricts the permission set to only allowing assets where my, where I've got, so I can, if I look at my permission sets, I can duplicate my reviewer set.

You might call this XYZ. And then in my asset filter, I can say the status, I only want them to see ready for review. So here I've set up my status equals ready for review.

[Transcript missing]

So I can do things like add my client to that filter. It will now show up in my permission set.

So the steps from that point is, I've got my permission set, it's restricting to only seeing those two or three assets. Then I create a group. I put my person from XYZ in that group, so when they log in as that user, they'll only be able to see two or three assets. Whatever has been marked as ready for review, where client equals XYZ.

And so once I've actually added my devices, added my users and permissions, It's then a matter of configuring the workflow. So setting up my watches and my scans. And a very simple example of that is if I wanted a folder, On my device here, I could set up a folder and send to play out, for example.

So then if I go to my system preference pane, I can really, I can easily set up a new automation. And so this is gonna be a file system watcher. And I can give it a name, you know, send to play out. And I can choose, you know, what device I'm going to watch. Well, I'm going to watch my content device and I'm going to look at the folder I created on there, which will be sent to Playout.

is the founder of Final Cut Server. He's the founder of the company that runs the So I can add a copy response and choose where it's going to go to. So it's going to go to my web server. It might go to a subfolder on my web server. And then I can choose my different transcode settings that are set up for that device.

So we can see here that we've got, you know, four responses to choose from in the pref pane, and I can go into much more detail in the...

[Transcript missing]

And finally, is the making of backups. Now with the backups, Final Cut Server will backup the database. And that's all. So you can actually set up point-in-time backups by saying backup now. or you can actually schedule a backup. So you can say, you know, back up at 1:00 AM every night.

And so once you've actually set up those backups, you're able to restore the database and the various preferences associated with it. But that's it. So a very important part of a backup is the ability to then backup the other material. So that might sort of be included in some system-wide backup or other bits and pieces, but when you need to restore, you can restore the is the founder of Final Cut Server. He's the founder of the backup using the restore mechanism here. And as long as all of the files which Final Cut Server was It'll be just fine, just as it was before you left. So if we go back to the slides.

[Transcript missing]

So, monitoring is very, very important, and there's a few different mechanisms to do that. So, there is the ability to look at a job page, and a, there's also, you can look at the logs in the client administration. and you can look at the system logs. In addition, if particular trans codes are failing, then you can look at a compressor batch monitor.

So, And then if you want to actually monitor particular failures, so say you wanted to be notified when a job to your very important playout server had failed. You can set up a subscription and an email to subscribe to failures going to that device, and send an email to an administrator.

So you can configure alerts on events, so a job failure, so forth and so on. And you should also use external system tools to actually monitor the health of the servers themselves. And there's plenty of options for doing that sort of things, but things like, you know, you want to know the disk usage, the bandwidth utilization, etc.

Now the job page, so this is an example of some jobs failing. Red is bad. And then if I double click on a red job, I can actually see what the error was. So here I've got an error submitting to Compressor. So then I can actually have a look in Compressor, and in this case, my Compressor cluster had failed, so I had to resolve that issue, and then I can continue my transcoding. And there's also the logs in the client administration which show you exactly what the server is doing at a given point in time.

Finally, just a few other considerations before I finish. So the first one is scalability. So how does the system scale? As we've seen, Final Cut Server has a database backend. As assets are added to the Final Cut Server, the resources required for the database increase. It's a matter of increasing both the memory and potentially the disk speed and the processor speed to actually be able to serve out more assets.

It's hard to give some exact figures on number of assets, but the system has been tested to go beyond millions of assets. But there are some considerations for search performance when you do get to that kind of level. So there's two different searches, you know, matches and contains. Matches quicker, contains slower. Because contains has to look across all permutations of all fields, whereas matches is much quicker.

You can definitely get to a lot of assets as long as the underlying infrastructure is able to support that. and then how many users. Again, it's very dependent on what the actual usage is. So what are the various usage patterns, or are people using edit in place, or are they actually downloading through the network? Support a large number of users on a system, but it depends on the underlying infrastructure. And then finally, how much throughput? Well, the great thing that we're seeing is compressor can be used to increase throughput. So as your throughput requirements increase, you can just add more nodes to a compressor cluster.

And just to talk a bit about hardware, so again, the system is very scalable, so you can install onto a Mac Pro, or I've got it running on a laptop over there. But for a medium to large deployment, you really want a nice, fast 8-core XRF. You definitely want rated storage, because you really don't want your proxies or your other versions to go away because a disk failed. and roughly, we're suggesting sort of about eight gigabytes for every quarter of a million assets. and as required, you can actually use technologies like Ethernet link aggregation to actually increase the possible bandwidth to your system.

Finally, I'll just talk a bit about one way of setting up the hardware which enables for a backup to be used. So, if I have Final Cut Server running on an XServe, And if I have all of my proxies and other variable media on an attached RAID unit or on an XAN, If I'm backing up the database, as we saw through the preference pane, to that attached RAID or that XAN, Then when that XServe fails, I can very simply install Final Cut Server on another XServe, You know, restore the database, and because all of the references are the same, Then my system is back up and running exactly as it was. So it's not high availability, but it's a very quick way to restore operations.

So that is, that's all I'm going to talk about. So as I've plugged a few times in the session so far, we've got to talk tomorrow, which is talking about integrating with Final Cut Server. And that's in Russian Hill at 10:30. And we're going to go into much more detail upon the integration with things like Podcast Producer, with Ruby on Rails, and with Spotlight. And then how to use things like the Execute Script Response, and much more detail upon the Watch and Response system.

and in relation to that, just in the session in here afterwards, there's actually a session on how to build Rails applications. So if you're going to come tomorrow, you've got any interest in Rails, I can highly recommend that one. And we've also got two labs. So we've got a lab right after this downstairs in the IT lab, and there's one tomorrow at 2 o'clock.

Now, I think that I have, just to summarize very quickly, we've got, so the key components to Final Cut Server that we saw, so we've got the media asset management side and the workflow automation side. We've got the flexible configuration options, and we've seen the possible steps that you can use in a deployment.