Error in pipeline when Deploying to Cloud Foundry
fifteentu opened this issue · comments
Getting this error during pipeline when using the Deploy To Cloud Foundry for watson-voice-bot:
Downloading app package...
Downloaded app package (1.6M)
-----> Python Buildpack version 1.7.10
-----> Supplying Python
ERROR Could not install python: no match found for 3.7.4 in [3.5.7 3.5.9 3.6.9 3.6.10 3.7.6 3.7.7 3.8.1 3.8.2]
Failed to compile droplet: Failed to run all supply scripts: exit status 14
I've found this re python-cloudpack:
“Could not install python: no match found for 3.7.4 in [ ... ]”
cloudfoundry/python-buildpack#158
Is there any quick hack/fix to make this Deploy To Cloud Foundry succeed?
Change version. in runtime.txt to python-3.7.7
I got this issue too and changing to a new version fixed it however the deploy fails with no particular error:
Waiting for app to start...
Start unsuccessful
TIP: use 'cf logs watson-voice-bot-xxxxx --recent' for more information
FAILED
Finished: FAILED
Any advice? Does the deploy have to be in a particular region? I deployed to London.
I'll fix the runtime version.
There is still a problem with the credentials. There is a known problem with the assistant creds role, but I'm not sure why there is a problem with STT/TTS. There is a workaround that works for both.
Go to the Cloud Foundry App -> Runtime -> Environment Variables and add the Assistant, STT, and TTS creds (APIKey and URL) using the same key/values described for use in the local runtime .env. We'll need to document this if we don't come up with an alternative.
Adding a screenshot here for now:
Similar issue re-occuring with release.txt
currently set to python-3.8.2
Supplying Python
**ERROR** Could not install python: no match found for 3.8.2 in [3.5.7 3.5.9 3.6.10 3.6.11 3.7.7 3.7.8 3.8.3 3.8.5]
Failed to compile droplet: Failed to run all supply scripts: exit status 14
I've removed the runtime.txt so we should be able to keep running on 3.x without breaking every time a minor version goes away. For the credentials workaround mentioned above, I created a separate issue. It is still an issue w/ workaround. Performance is still an issue for me too.