Re:Gitlab cancels plan on tracking user behavior on GitLab.com
by badrequest on Tuesday October 29, 2019 @04:26PM
(#21391593)
It's a Rails app trying to load hundreds of different bits of data (emojis, comments, votes on comments, etc.), I'd honestly be more surprised if it were quick.
Re:Gitlab cancels plan on tracking user behavior on GitLab.com
by thenewnewguy on Wednesday October 30, 2019 @12:01PM
(#21399147)
GitHub loads similar data and is also written in rails, but long GitHub PRs/issues are significantly faster to load.
Re:Gitlab cancels plan on tracking user behavior on GitLab.com
by dijit on Tuesday October 29, 2019 @04:27PM
(#21391601)
It’s probably hammering the database right now.
These kind of things, by nature, are difficult to cache because the balance between “fresh” new content and how often they are accessed leans heavily towards having it served from the backend directly.
Re:Gitlab cancels plan on tracking user behavior on GitLab.com
by ericmcer on Tuesday October 29, 2019 @05:59PM
(#21392553)
Resources are cached, lightweight JSON that dictates how to lay those resources out is fresh, doesn’t seem tooooo difficult
Re:Gitlab cancels plan on tracking user behavior on GitLab.com
by swagonomixxx on Tuesday October 29, 2019 @04:28PM
(#21391612)
Ruby on Rails backend, I think.
Re:Gitlab cancels plan on tracking user behavior on GitLab.com
by heartbreak on Tuesday October 29, 2019 @05:11PM
(#21392073)
Github is also RoR backend, and it's not this slow.
https://imgur.com/a/uxaU0h8
How is GitLab still this slow?