Shared cache on NFS Introduced #455
Conversation
| link data files from cache to your workspace. It enables symlinks to avoid | ||
| copying large files. | ||
|
|
||
| `cache.protected true` - to make links `read only` so that we you don't corrupt |
There was a problem hiding this comment.
again explain it better that since we are going to use symlinks in this case between cache and workspace (since they are located on different file systems) it important to protect files so that we don't corrupt the cache accidentally. Mention that dvc unprotect should be used in this case, link to the https://dvc.org/doc/user-guide/update-tracked-file
There was a problem hiding this comment.
@shcheklein we need dvc unprotect only when we are writing to NFS directly?
There was a problem hiding this comment.
we need to run dvc unprotect in the client's workspace if we want to edit/rewrite the file that is under DVC control.
| Now, add first version of the dataset into the DVC cache (this is done once for | ||
| a dataset). | ||
|
|
||
| ```dvc |
There was a problem hiding this comment.
let's simplify all this workflow. Let's just ask users to SSH into NFS serve machine, do git clone .../project. Move data into project and run dvc add, git commit, git push, (dvc push optional) after that. All the stuff below can be adjusted a bit.
There was a problem hiding this comment.
@shcheklein i think it will just confuse the user.
There was a problem hiding this comment.
cp -r . /project/ is very confusing also. I would say we need to explain the motivation here - we want to avoid copying existing data to a client machine to take it under DVC control.
I also, think git clone protocol is a standard way to collaborate and update different requirements. It's better to do this from the NFS server machine. It'll emphasize that NFS takes care about data.
|
@shcheklein please review this. |
| possible and have a workspace restoration/switching speed as instant as | ||
| `git checkout` for your code. | ||
|
|
||
| With large data files it is better to set the cache directory to external NFS. |
There was a problem hiding this comment.
minor: we use it's - it's less formal
| `git checkout` for your code. | ||
|
|
||
| With large data files it is better to set the cache directory to external NFS. | ||
| Not only just it will cache the data faster but also version the data. Suppose, |
There was a problem hiding this comment.
cache faster - I'm not sure I understand this
| of a complete dataset. With `cache directory` set to `NFS server` you would | ||
| avoid copying large files from NFS server to the machine and DVC will manage the | ||
| links from the workspace to cache. | ||
|
|
There was a problem hiding this comment.
the paragraph above is good but feels repetitive to the first paragraph in the document and has minor problems. What information you are trying to deliver here? Can you summarize it here in the comments, please? And we'll see how can we improve the text.
| @@ -0,0 +1,148 @@ | |||
| # Shared Storage on NFS | |||
|
|
|||
| In the modern software development environment, teams are working together on | |||
There was a problem hiding this comment.
software development -> machine learning
(I even agree that it's software engineering, but it's bette to delineate them for now)
| @@ -0,0 +1,148 @@ | |||
| # Shared Storage on NFS | |||
There was a problem hiding this comment.
I want to rename it ... it's not about NFS only. It's about any network attached storages. We can do something like:
Share Storage on NAS (NFS)
|
|
||
| In the modern software development environment, teams are working together on | ||
| same dataset to get the results. It became necessary that data is accessible and | ||
| every team member has a same updated dataset. NFS (Network File System) storage |
There was a problem hiding this comment.
NAS (NFS is one common example) is widely ...
There was a problem hiding this comment.
we can also mention something like: "Here we would like to show you how to setup a shared cache on NFS, but the same idea applies to any other NAS"
| team member is using the same cache location. | ||
|
|
||
| After configuring NFS on both server and client side. Let's create an export | ||
| directory on server side where all data will be stored. |
There was a problem hiding this comment.
it's better to use : when you have a code block you are writing about in the sentence
| From `/mnt/dataset/` you will be able to access `/storage` directory present in | ||
| host server from your local machine. | ||
|
|
||
| ## Configuring Cache location |
| Next, you can easily get this appear in your workspace by: | ||
|
|
||
| ```dvc | ||
| $ cd /home/user/project/ |
There was a problem hiding this comment.
before the project path was /project
| information on `.dvc` file format, visit | ||
| [here](/doc/user-guide/dvc-file-format). | ||
|
|
||
| `data` directory will now be a symbolic link to the NFS storage. |
There was a problem hiding this comment.
worth writing something similar to the last paragraph in the introduction to reiterate on why links are important, worth showing an output of the ls -a
|
@ryokugyu any updates on this? :) it's almost done as far as I can tell, would be great to get it merged. |
@shcheklein will work on it. Sorry for the delay! |
|
I think that the "Mounted DVC Storage" (which is explained on this interactive example: https://katacoda.com/dvc/courses/examples/mounted-storage) is more general than just NFS and it deprecates this one. |
my concern that it's very specific because of SSHFS and it's not emphasized enough that NFS, NAS (whatever else?) is covered
don't think so. Especially the way interactive tutorials are made - they are extremely dry and do not explain motivation very well, do not explain what is happening behind the scene and what commands are doing. |
In this case it is just an interactive example (not a tutorial) and it is referenced from a User Guide page: https://dvc-org-pr-784.herokuapp.com/doc/user-guide/data-sharing/mounted-storage |
kk. It just in you initial comment you mentioned only the interactive tutorial and hadn't had enough time to see the UG changes. Will get back to this one when I have time to read the epic PR :) |
|
I think it is still relevant. Unfortunately, no easy way to reopen it now. |
fix #103