-
Notifications
You must be signed in to change notification settings - Fork 2.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
amtool: Detect version drift and warn users #2672
Conversation
This change detects the alertmanager version when initiating the client. It ignores most errors since I expect amtool to fail later. If amtool is not compiled with proper version, we do not do anything either. We use MajorMinor for now as we have not reach 1.0, but we still allow the bugfix version number (Z in x.y.Z) to differ. Signed-off-by: Julien Pivotto <roidelapluie@inuits.eu>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Approach LGTM, some quick questions on whenever we should have a way to make this optional.
@@ -93,7 +95,19 @@ func NewAlertmanagerClient(amURL *url.URL) *client.Alertmanager { | |||
cr.DefaultAuthentication = clientruntime.BasicAuth(amURL.User.Username(), password) | |||
} | |||
|
|||
return client.New(cr, strfmt.Default) | |||
c := client.New(cr, strfmt.Default) | |||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Given this is now executed for every command, can we an option to skip it?
Something like:
var skipVersionCheck bool
app.Flag("skip.version-check", "Skip version checks when executing commands.").BoolVar(&skipVersionCheck)
// Then, we can do something like
if skipVersionCheck {
return c
}
|
||
status, err := c.General.GetStatus(nil) | ||
if err != nil || status.Payload.VersionInfo == nil || version.Version == "" { | ||
// We can not get version info, or we do not know our own version. Let amtool continue. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Now that we have an option to cancel the check, I don't think we should fail silently. How do you feel about logging that we were unable to compute the version and tell the user that should upgrade/downgrade?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If we can not run the check, it is because alertmanager does not work at all, and the error will be printed in the actual call later on.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I was thinking more in the case of Grafana but this is not a concern for us here to be honest.
return c | ||
} | ||
|
||
if semver.MajorMinor("v"+*status.Payload.VersionInfo.Version) != semver.MajorMinor("v"+version.Version) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is quite nice 🎉
Signed-off-by: Julien Pivotto <roidelapluie@inuits.eu>
updated |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM.
$ ./amtool --alertmanager.url=http://localhost:9093 config show
Warning: amtool version (0.22.2) and alertmanager version (0.21.0) are different.
# and then
$ ./amtool --alertmanager.url=http://localhost:9093 config show --no-version-check
Both work as expected.
This change detects the alertmanager version when initiating the client.
It ignores most errors since I expect amtool to fail later.
If amtool is not compiled with proper version, we do not do anything
either.
We use MajorMinor for now as we have not reach 1.0, but we still allow
the bugfix version number (Z in x.y.Z) to differ.
Signed-off-by: Julien Pivotto roidelapluie@inuits.eu