Hi
Im setting up the pipeline for a bunch of Episodes, we have certain episodes stored on different file server mounts for example:
/mnt/fs1
- 101
- 102
- 103
/mnt/fs2
- 104
- 105
- 106
The Episode is at the root of the Volume, within the pipeline I would like to be able to reference each episode/scene/shot without having to put them all under a single project folder.
Does anyone have any pointers on how I can achieve this?
Thanks
Dan.
Hi @Daniel_Marshall,
Welcome to the forums! I do believe that this is possible, likely it would be a 2 step process, however I donāt quite know the details of the second step, Iāll need to ask one of my colleagues to chime in there.
The first step would be to set up your config with a second root, (a multi-root config) we have a doc on our developer site with a setup workflow here
As for the second step, as this is having different roots for a single entity rather than different entities, we need to modify episodes in a way that enables them to be recognised as either primary or secondary. As noted above Iām a bit scant on details here, but Iāll get back to you when I find out more.
-David
Hi David
Thanks for this, I had a look through that article.
Im thinking I may be able to couple this with a custom field which defines which āblockā an episode is stored on, so the query syntax would know which block and adjust the root to matchā¦ how that
works in practice im not sureā¦ but ill have a play round.
Cheers
Dan.
Great stuff, let us know how you get on if you solve before I have a chance to get back to you!
Hi David
OK so I got it splitting the data between volumesā¦ couple of bits im stuck on:
I can only create objects of type āprojectā in the root of a folder, at the moment the root of block1 is my episodeā¦ ideally id like to keep
it this way if possible.
I need to understand how I can scope the filters better if you have any pointers. Ultimately I think I need to scope the filter so that it
can take a range of valuesā¦ i.e Episode 101, 102, 103 -> Block1, Episode 104,105,106, Block2
If anyone can point me in the right direction that would be great.
Thanks
Daniel.
Hi @Daniel_Marshall,
Sorry for the long reply delay, I wasnāt 100% on the response I received and easter holidays happened 
Unfortunately itās a limitation of tk that your schema needs to start with a project folder but we thought of a way to accomplish this;
episode/ folder represent TWO folders in your filesystem, mount/episode ā you can get the values for both from fields on the episode entity: https://support.shotgunsoftware.com/hc/en-us/articles/219039868-Integrations-File-System-Reference#Multiple%20folders
There is one possible caveat: are the mounts represented the same way on all OSes/do you intend on using this in windows and OSX?
-David
Hi David
Ok thatās no worries thanks for coming back to me, I was expecting that to be the case, its not many times people would ever want to create everything outside of a project.
Actually changing the mountpoint probably wont be an issue, we would just end up with:
/mnt/<Project_Name>/block1
Ive made a custom field within Shotgun at the Episode level called āsg_block_storage_locationā and then just give that values such as āblock1ā, āblock2ā etc, I then pull that into the episode template
with { sg_block_storage_location}/{code} and that gives me something like āblock1/404ā which I think was what you was meaning below and works for me, never thought of doing it that way round.
The filesystem in use is AWS FSx for Linux, so you cant map any windows clients to itā¦ Ultimately though they would all see the root mountpoint the same anyways, so not sure if that would be an issue.
Just one thing I wanted to confirmā¦ Im sure this is a NO, but Does Shotgun overwrite any existing folders with the same name when doing folder creation? Obviously ive a whole lot of pre-existing
paths which are on those block1 devices and I donāt really want to erase everything.
Thanks
Dan,
Hey @Daniel_Marshall,
Toolkit never deletes or overwrites anything on your filesystem so you are good there, it will check the filesystem and create a filesystem location entity in SG, but you can run create_folders as much as you want, it wonāt overwrite or delete.
-David