[Documentation] Install guide demonstrates how to copy a pipeline from Curoverse Cloud
Use case: an administrator just went through the install guide, and now they have a nice, shiny new cluster. What do they do with it?Write a new page at the end of the install guide that describes:
- How to browse public projects on cloud.curoverse.com to find an interesting pipeline
- How to copy the selected pipeline project to your own cluster using arv-copy
You don't need to introduce arv-copy itself. Instead, build on top of the user guide written in #5288 to describe how to copy the chosen project or pipeline template specifically.
#3 Updated by Radhika Chippada almost 5 years ago
I noticed an issue as I was writing this document. It is:
4:44 The user needs current_token to be able to use arv-copy.
We can hard code current anonymous token directly into the documentation.
Which of course is bad if we need to replace it in the future (for security purposes)
I think the user should be able to get it from Workbench. Do we want to add a “Current token” menu option to the Help dropdown when the system is being accessed by an anonymous user?
4:52 I don't think there's any reason not to put it right on the doc page...
4:55  That's no more public than Workbench-when-not-logged-in,  If we change it, we can update the docs just as easily as updating the Workbench config?
4:55 I suppose we might forget to update it in the docs, though.
4:55 I am actually testing if it will work with anon token though.
4:56 I do not think the recursive copying, that requires a repository, would work anyways.
4:56 I wonder whether it makes sense for the API to give out the anon token in the discovery doc.
4:57 (The reason we don't make it "magical" on the API side is to make sure clients can control which token they're using, and keep the auth paths relatively simple. But I don't know if there's any reason we couldn't give one out in the discovery doc.)
4:58 Right, I just mean the SDK & arv-copy could learn how to get an anonymous token if you say --anonymous or something.
4:59 ah. that sounds good.
4:59 Seems much easier to put it in the docs
4:59 Less steps for the user.
4:59 for now i will hard code.
4:59 and test if it works or not with anon token.
4:59 Right. If it doesn't we should fix that. :)
5:00 ok. sounds good. thanks.
#7 Updated by Bryan Cosca almost 5 years ago
- Story points deleted (
"Let's access one such public pipeline and copy to your cluster so" > "Let's access a public pipeline and copy it to your cluster, so"
"use it to create the source qr1hi.conf configuration file." > "use it to create the source qr1hi.conf and dst_cluster.conf configuration file."
I think before you say "Configure the source and destination clusters as described by the Using arv-copy tutorial", you should say something like if you have a Curoverse cloud beta account, you can follow the instructions. If you do not however, please see the "Configuring the source using anonymous access token" part of the documentation. Otherwise, they could get lost in the using arv-copy documentation without realizing they could just use the anonymous access token. Then remove the "On the other hand..." Then I might point again to the arv-copy documentation to setup the dst_cluster config.
(I realize I might make this more confusing but I think the flow is a little weird right now because you tell them to configure the source and destination clusters using the tutorial, and if they go do that without reading the entire paragraph, they will get stuck)
#9 Updated by Bryan Cosca almost 5 years ago
"Configure source and destination while using anonymous access token for source" -> "Configuring source and destination setup files using anonymous access token"
"Configuring using your access tokens" -> "Configuring source and destination setup files using personal access token"