Network Setup - Basic Questions

[mtjj]

Hi, currently trialling IMatch with a view to moving away from adobe products, and I'm struggling to get my head around the basic setup needed. I hope no one minds if I ask some questions, I've read through a lot of the help files and various online articles, but I think sometimes they are based on slightly different use cases from what I am trying to do, either that or I'm a bit slow today

This will be a fairly long post to give context, but I'll highlight the questions so people are not obliged to read it all 

Background / High Level Use Case

My photo workflow consists of three steps:
1 - initial import / tagging / selection of keepers / etc. from RAW files - e.g. import 100 photos, convert to DNG, adjust time, apply tags, apply GPS coordinates, select 20 photos to process. Currently use Lightroom
2 - actual processing and generation of processed JPGs. Currently use lightroom and/or Photoshop and/or other programs.
3 - viewing / management of processed library - e.g. browse all photos taken at x location, make an album, save a set of reduced sized JPGs for web sharing with watermarks and set of full size JPGs with borders for my own use. Currently use windows explorer and a bunch of different tools in various workflows that don't really make much sense....

Initially I want to use IMatch for step 3, then if that works well, for step 1 also. The primary goal right now is to have something I can use for browsing / filtering / reviewing photos that have already been processed, and exporting modified copies for various purposes. Ideally I want to use IMatch Anywhere so that I can have a client / server model with all my processed photos sitting on a home server that I (and my family) can access from other devices, although I know the functionality of IMatch Anywhere is fairly limited at the moment.

My setup consists of:
"Workstation" - Primary image editing machine (Windows 10 Pro). I do steps 1 and 2 on this PC, and then archive the RAW files onto external hard drives. I login as user MTJJ. This computer is not always switched on.
"Server" - This is a home "server" (Windows 10 Pro) that has all my media on it, on a series of connected (USB) drives. The processed JPGs go on to a drive attached to this computer. I login as user MTJJ_Server. This computer is always on and connected to the network and internet.

The drive with the photos on it (identified as drive "W" on the Server), is shared on the Server to user MTJJ, so that from the Workstation I can access it over the network, and I also map it as "W" on the Workstation (so that I can easily saved processed JPGs to it).

(I find it cheaper and much easier to just buy a bunch of external drives and pug them to the server than it would be to buy a separate NAS. I'm aware there are drawbacks to this approach, but I'm not likely to change this setup any time soon.)

Question 1
Where should I install IMatch, where should I install IMatch Anywhere, and where should I store the database?
I'm fairly sure that IMatch Anywhere needs to go on the Server, along with the database. But can I install IMatch on the workstation and access the DB on the Server? Because I initially plan to use IMatch only to work with already-processed files I don't expect to make many changes other than exporting copies and building collections, so can I use a single DB and have IMA access it in read only mode? Or is it better to have two copies of the DB, one for IMatch and one for IMA, and if so, how do I keep them in sync? (I am trying to avoid anything that requires me to remotely log into the server, if possible.)

Question 2
How do I then map / relocate the drives?
This is dependent on question 1, but if the same DB is sitting on the Server, accessible to IMatch Anywhere on the Server and IMatch on the Workstation, which path do I need to map (e.g. W:/Photos to //Server/photos ?)

Question 2
How do I setup users / accounts / access in IMA?
I tried to setup a trial of IMA on the Workstation and this is the bit I really got stuck on 
I installed the demo and the service started, but couldn't access any pictures, because I hadn't mapped W to //Server/... So I did this, and tried to change the user settings, but now something has gone wrong and the service won't start. (Errors out with a login error 1069). I tried reinstalling, deleting the config xml so it would re-create it, but whatever I change, I can't get the service to start again.
There are 3 different different "users" on the IMA configuration and I don't understand what is what:
- Main screen - database user. This is just my windows login name? I'm not sure what the purpose of this is?
- Advanced Options - Service User Account. I don't understand the help file here, is this needed if the data (photos) sit on a separate computer / NAS, or only if the database itself is on a separate computer? I anyway entered my Windows login information here too, but it doesn't seem to make a difference.
- Enable Authentication - users. I don't think this is necessary to get the service running, correct? I tried setting this up as well, but it didn't seem to make a difference.

Right now, whatever I change, I get the same 1069 error. Also, the service log is not updating, not sure why...

Would be most grateful for any advice - or if there is some setup guide I am missing, please let me know!

[Mario]

Hi, welcome to the community.
Sorry for the late reply but apparently this massive amount of text has frightened your fellow users. Or it's just the time of the year 

Question 1:

IMatch is basically a database application with a graphical user interface. It's performance depends heavily on the speed of the disk on which the database is stored. I recommend storing the database on the fastest disk and on the same machine you run IMatch on.

Pulling images over the network is 10 to 1000 times slower than accessing images from a local disk. If you install IMatch on your workstation but your files are on your server, you will get slow performance when IMatch is ingesting the files. Afterwards IMatch only accesses the remote files when it needs to create or update a local cache image. Or when you write-back metadata.

Tip: For the initial indexing process (guessing that you have hundreds of thousands of files to add), I would suggest that you install IMatch on your workstation and on your server (I allow multiple installations if the licensed user is the only user).

Create your database on your server and index all your files there. Use UNC shares so you don't need to relocate later. This gives you the best possible performance for the initial indexing.
After creating the database on the server machine at max. speed, move the database to your workstation and use it there with IMatch.

In your scenario I recommend that you use IMatch Anywhere with a publishing workflow. You work with IMatch on your desktop. To make the current version of your database available via IMatch Anywhere, you copy the current version of the database to the server. This can be automated (shutting down IMWS, copy the database, restart IMWS) easily using a small PowerShell script. You can run that manually of automated.

Question 2:

If you use UNC shares you don't need to relocate. They work in your entire network.
IMatch has also "Portability Options" which automatically relocate drives or folders in a database when you move it to another computer (see the IMatch help for details).

Question 3:

I installed the demo and the service started, but couldn't access any pictures, because I hadn't mapped W to //Server/...

IMWS is by default installed using the Local Service user account. This means it can access all resources on the local system, but no images on remote computers. See the IMA documentation for details and how to deal with that.

Main screen - database user.

Please click on help or press <F1> when the IMatch Anywhere Controller application is running. The documentation explains the meaning of this.

Advanced Options - Service User Account

See my comment above (Question 3) and also the documentation. This is the user account under which you run IMWS to allow it to access remote resources (your files on the NAS/Server).

Enable Authentication - users. I don't think this is necessary

No. This is only used when you allow more than one user access to IMatch Anywhere and you want to apply restrictions for what these users can see or do with IMatch Anywhere. If you use IMA alone or only in your home network, you don't need user management.

[mtjj]

Hi, welcome to the community.
Sorry for the late reply but apparently this massive amount of text has frightened your fellow users. Or it's just the time of the year 

Thanks, it's New Year's Day - actually I wasn't expecting any response at all for a few days, thank you for taking time off to post I've worked with a lot of developers and I've seen them get annoyed when people just post a one line "dude, this doesn't work" and then expect them to figure out the context - I may have gone too far the other way with too much info!

Question 1:

IMatch is basically a database application with a graphical user interface. It's performance depends heavily on the speed of the disk on which the database is stored. I recommend storing the database on the fastest disk and on the same machine you run IMatch on.

Pulling images over the network is 10 to 1000 times slower than accessing images from a local disk. If you install IMatch on your workstation but your files are on your server, you will get slow performance when IMatch is ingesting the files. Afterwards IMatch only accesses the remote files when it needs to create or update a local cache image. Or when you write-back metadata.

Tip: For the initial indexing process (guessing that you have hundreds of thousands of files to add), I would suggest that you install IMatch on your workstation and on your server (I allow multiple installations if the licensed user is the only user).

Create your database on your server and index all your files there. Use UNC shares so you don't need to relocate later. This gives you the best possible performance for the initial indexing.
After creating the database on the server machine at max. speed, move the database to your workstation and use it there with IMatch.

Got it - that makes perfect sense. I think if I use UNC shares from the start it also avoids the knot I got into below as well

...(snip)...

Advanced Options - Service User Account

See my comment above (Question 3) and also the documentation. This is the user account under which you run IMWS to allow it to access remote resources (your files on the NAS/Server).

... and I think this becomes a lot simpler if I'm running IMWS on the server, because it's always the same user....

Enable Authentication - users. I don't think this is necessary

No. This is only used when you allow more than one user access to IMatch Anywhere and you want to apply restrictions for what these users can see or do with IMatch Anywhere. If you use IMA alone or only in your home network, you don't need user management.

Just checking, I can leave this for now and setup other users later, right? Or is there a minimum I need to setup at the start to avoid creating problems later?

Many thanks!!

[Mario]

You don't need to configure user accounts. Leave authentication off and all is fine. Enable as needed later.

[mtjj]

I set this up today on the server and it all worked fine - with the DB on the server and UNC paths from the start there were no issues. Just one (very) minor point, I tried to access IMA from the workstation with [server-name]:8801 and got a 403 error, it took a minute or two to realize it needed to be [server-name]:8801/imatchviewer/ - you might want to do a redirect. The other media server programs I have (emby for music/video and calibre for ebooks) just work on the port number.

Thanks again, just need to work out the DB update / replacement process, but I think IMatch will do what I need.

[thrinn]

it took a minute or two to realize it needed to be [server-name]:8801/imatchviewer/

That's because imatchviewer is only one of possibly several endpoints (or apps). For example, the IMatch Web Services Controller already displays a second one, namely http://<YOUR-PC>:8081/info. I assume the other media server programs you mention have exactly one entry point?
It is true that at the moment imatchviewer is the only app I use with IMA, but this may change in the future. Therefore, having the app oder endpoint name as part of the URL makes sense for me. But I have to admin, I also forgot entering the imatchviewer part multiple times, because I simply neglected that the correct URL is shown in the IMA Web Services Controller...

[Mario]

That's right. IMA is not tied to IMatch the WebViewer client app. You could also install IMatch apps on the IMWS server and run them. Or PHP, Python scripts, or anything. I did not want to complicate things by hard-coded redirects.

The WebService Controller displays the correct URL for IMatch WebViewer. And also a test URL for the /info endpoint.
The correct URLs are also copied into the clipboard when you use the "Copy Info" button.

Only when you type the URL from hand into the browser address bar and you leave out the /IMatchViewer/ part you run into problems. But, at least according to the support cases, not many users do that.