You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
Currently, a log warning is produced when any failure during span export occurs.
In looking at the JaegerGrpcSpanExporter in particular, I was curious how it's possible to know whether a failure to export is a transient failure vs a more serious problem, such as an endpoint that can not be connected to because the URI is wrong, even if it's a valid URI.
Describe the solution you'd like
In the situation of an endpoint that can't be connected to, maybe it's possible to track such a connection issue and log an error instead of a warning if that same error occurs a certain number of times, and then stop logging if the failure continues.
I appreciate it needs to handle transient failures and log those at a warning level, but for a persistent failure, something more than a log warning would be preferred.
The text was updated successfully, but these errors were encountered:
@kenfinnigan do you have any ideas on how to actually implement this feature? Would you be willing to write the code for one of the exporters to support this?
Is your feature request related to a problem? Please describe.
Currently, a log warning is produced when any failure during span export occurs.
In looking at the
JaegerGrpcSpanExporter
in particular, I was curious how it's possible to know whether a failure to export is a transient failure vs a more serious problem, such as anendpoint
that can not be connected to because the URI is wrong, even if it's a valid URI.Describe the solution you'd like
In the situation of an endpoint that can't be connected to, maybe it's possible to track such a connection issue and log an error instead of a warning if that same error occurs a certain number of times, and then stop logging if the failure continues.
I appreciate it needs to handle transient failures and log those at a warning level, but for a persistent failure, something more than a log warning would be preferred.
The text was updated successfully, but these errors were encountered: