Hello, a colleague (who's not available right now ...
# general
d
Hello, a colleague (who's not available right now so I can't ask much) had an issue with his internet connectivity while doing a pulumi up. The state is now in a bit of a 'state'... A refresh is not helping and pulumi thinks that it has to create almost the entire stack from scratch. I am tempted to pick a json file from the .pulumi/history folder and save it into ./pulumi/stacks but I don't want to mess things further. Any idea how to get out of this mess? Thanks
clearly something wrong with the size of files
that seemed to have worked.. really nice the history command which I wasn't aware of, I could tell the last failed stack without going into CI.. it would be awesome to have a restore command which picsk up the last successful apply ...
l
Which history command?
r
I had faced similar issue , I recovered it by manually syncing the older correct state in the s3 bucket having the stack
d
@little-cartoon-10569 you can run
pulumi stack history
it will give you a log such as
Copy code
UpdateKind: update
Status: succeeded
Message:
+0-0~7 131 Updated 4 days ago took 3m51s
    ci.system: AWS CodeBuild
    exec.kind: cli
it always says version: 0 though , it would be good if anyone knows how to set that , perhaps with a git hash?
e
@damp-school-17708 sorry to ping you, but I’m facing the same issue. How did you actually made a “rollback” to a given checkpoint in the state file history folder? Much appreciated any help, worried to make a wrong command and loose even more.
Just a “copy and past” of files? If so, which ones, if you could tell me.. thanks!
d
yeah I basically downloaded the checkpoint, renamed it, and overwritten the normal file
👍 1
keep a copy as always in these cases of the orginal
😅
e
Thanks for the response! I did the renaming and I’m back 🙌