ddev: Allow ddev config pantheon to specify an explicit site hash to override what the API may find (or fail to find)
Is your feature request related to a problem? Please describe.
I’m an admin for UC Berkeley – we have something like 1500 campus sites in our Pantheon org. When I run ddev config pantheon
sometimes it can’t find the site (I get a “could not find a pantheon site named sitename” message.). Passing a projectname
doesn’t appear to help find the site either.
I tried running ddev config
instead. Then I changed the following in config.yaml: provider: pantheon
. Then I copied an import.yaml file from a different site that Pantheon could find, and changed site
to the correct Pantheon machine name.
Unfortunately when I try running ddev pull
(or ddev start
) it still doesn’t find the site – it appears to still be searching Pantheon as opposed to using my local yaml changes.
Describe the solution you’d like
It appears that ddev’s projectname
is searching for and trying to match the sitename string, not the hash.
It would be helpful if we could specify the site via ddev config pantheon --projectname string
or something like ddev config pantheon --projectID hashnum
since we can find the Pantheon site ID via terminus site:info sitename
. I’m hoping that would avoid having ddev try to look up the site by name each time and failing if it exceeds the Pantheon API limit.
Describe alternatives you’ve considered
We’re able to use ddev config
and then manually import db/files, but we’re looking to use a solution that provides better Pantheon integration.
About this issue
- Original URL
- State: closed
- Created 6 years ago
- Comments: 15 (9 by maintainers)
Commits related to this issue
- Rework pantheon integration to use terminus inside web container, fixes #1731, fixes #1232 (#1827) — committed to rfay/ddev by rfay 5 years ago
Yay, thanks, I added a commit suggesting the name-match problem when that happens.
@ucbklu @gchaix @frazras I sure would appreciate if you could test #1827, it uses terminus, pantheon’s own interface (inside a container) to interact with Pantheon, so I’m hoping that that could fix the problem in this issue. But since there seems to be no way to recreate it without privileged access to one or your accounts, I’ll have to ask you to see what you can figure out. Artifacts are provided over there in #1827
Thanks @gchaix - I do have a PR underway (but neglected) to completely rewrite Pantheon support in terms of terminus execution, which ought to solve this. https://github.com/drud/ddev/pull/1827. I certainly do hope to get it done in the next month or two, I hope you’ll be willing to review/try it out. Because it’s great to have a guinea pig with an unreproducible problem.