Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Caching of object artifacts and config #76

Open
sbcgua opened this issue Jul 2, 2020 · 3 comments
Open

Caching of object artifacts and config #76

sbcgua opened this issue Jul 2, 2020 · 3 comments

Comments

@sbcgua
Copy link
Collaborator

sbcgua commented Jul 2, 2020

Ideas from #72

  • Problem 1: the config is sent with each request
  • Problem 2: dependencies are sent in each request

Ideas:

  • Addressing just 1: Supposing that abaplint backend would be most probably private, we can store configs as key/value. Where key would be repo url (which is uniq enough) and value - the config. The assignment can be done from UI. So adding a new package at UI side would ping server side to check presence of the config and submit it if missing or changed
  • Addressing both: implement sessions. On SCI start, it requests (or generates UUID) a session key, and then server can store the cache for some period of time. Including config and dependencies ! Does SCI keeps same abap session between object to have singletons ? This can be a bit tricky but totally implementable

Maybe both can be implemented, because the actually solve a bit different nuances. E.g. storing configs may be useful to define a global corporate config like descussed in #72

@larshp @g-back

@larshp
Copy link
Member

larshp commented Jul 2, 2020

Problem 1: practically, how much time in percent of the total time would be possible to save by not transferring the config file? just a guess?

repo URL: I also see a use-case for this repository, to not have the actual code being checked in git, just have it in the abap system, not really using abapGit for anything

@sbcgua
Copy link
Collaborator Author

sbcgua commented Jul 2, 2020

Problem 1: practically, how much time in percent of the total time would be possible to save by not transferring the config file? just a guess?

This is a fair point. I think cannot be answered without measuring. First thoughts why this would be good: unified data shape, potential possibility to cache parsed config (though it is more towards sessions in the question 2). In my opinion sessions would give more impact, so I'd better target p2.

repo URL: I also see a use-case for this repository, to not have the actual code being checked in git, just have it in the abap system, not really using abapGit for anything

Hmm, local://our-secret-corporate-packge ? or local://our-secret-corporate-packge@hardwarekey+sysid

@sbcgua
Copy link
Collaborator Author

sbcgua commented Oct 4, 2020

Seeing how much time takes deps generation I'm not sure if server caching will do any better :(
I'm currently for pre-caching to a repo or otherwise but not during runtime

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Development

No branches or pull requests

2 participants