helm: Helm doesn't cleanup on unsuccessful deployments
After encountering an error during install/upgrade helm doesn’t clean after itself (that is operation isn’t atomic), which forces me to clean manually each time during testing. Also, that deployment name isn’t registered, thus I cannot use helm delete --purge <name>
~ helm version
Client: &version.Version{SemVer:"v2.7.2", GitCommit:"8478fb4fc723885b155c924d1c8c410b7a9444e6", GitTreeState:"clean"}
Server: &version.Version{SemVer:"v2.7.2", GitCommit:"8478fb4fc723885b155c924d1c8c410b7a9444e6", GitTreeState:"clean"}
About this issue
- Original URL
- State: closed
- Created 6 years ago
- Reactions: 10
- Comments: 16 (4 by maintainers)
Commits related to this issue
- [#3338] Rename "safe" to "atomic" which gives more clear understanding, what it does Signed-off-by: Alexander Nesterenko <nestorf250@gmail.com> — committed to cleac/helm by cleac 5 years ago
- [#3338] Rename "safe" to "atomic" which gives more clear understanding, what it does Signed-off-by: Alexander Nesterenko <nestorf250@gmail.com> — committed to cleac/helm by cleac 5 years ago
- [#3338] Rename "safe" to "atomic" which gives more clear understanding, what it does Signed-off-by: Alexander Nesterenko <nestorf250@gmail.com> — committed to cleac/helm by cleac 5 years ago
- [#3338] Rename "safe" to "atomic" which gives more clear understanding, what it does Signed-off-by: Alexander Nesterenko <nestorf250@gmail.com> — committed to cleac/helm by cleac 5 years ago
- Merge pull request #5143 from cleac/master Fixes #3338: Implement "atomic" upgrade/install — committed to helm/helm by thomastaylor312 5 years ago
- Remove #ready? method in favor of #available? Given https://github.com/helm/helm/issues/3338, I think that we should exclude applications that are :update_errored, :updating as well. — committed to tigefa4u/gitlabhq by deleted user 5 years ago
- Remove #ready? method in favor of #available? Given https://github.com/helm/helm/issues/3338, I think that we should exclude applications that are :update_errored, :updating as well. — committed to tigefa4u/gitlabhq by deleted user 5 years ago
- [#3338] Rename "safe" to "atomic" which gives more clear understanding, what it does Signed-off-by: Alexander Nesterenko <nestorf250@gmail.com> Signed-off-by: Kevin Labesse <kevin@labesse.me> — committed to eraac/helm by cleac 5 years ago
- Merge pull request #5143 from cleac/master Fixes #3338: Implement "atomic" upgrade/install Signed-off-by: Kevin Labesse <kevin@labesse.me> — committed to eraac/helm by thomastaylor312 5 years ago
this is horrible…getting pissed off…most of the time i have delete and re install…thats not a trivial solution. Anyone have any other approach
This issue is frustrating, Why do helm charts gets installed partially , when the whole chart isn’t perfect? Any work around commands for this?
ERR : Error: release t1 failed: storageclasses.storage.k8s.io “xxxxx” already exists
I hit this error frequently. An install fails and then I make some changes to my config and try to install again but hit errors complaining about how many pods/configmaps etc already exist. When you run helm list it doesn’t list any failed or successful deployments so then I have to go and cleanup everything manually. If an install fails it should definitely rollback the changes it has already made not leave a half installed helm chart.
@dpepper is
helm list -aany better?