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
Charon has been designed from the beginning with a no-disk-write approach: aside for the DKG and create cluster, Charon never writes anything on disk.
The container-exclusive deployment strategy leaves the burden of log retention to the user: since on container restart logs are lost, the user needs to devise a plan to persist them somewhere in order to audit and eventually find and report bugs to us.
🛠️ Proposed solution
Approved design doc: link
Core team consensus on the proposed solution
I propose we consider the implementation of a on-disk log retention strategy, where Charon writes logs on the already existing facilities (Loki, stdout) as well as files - plural here because log files need to be periodically rotated.
Zap provides multiple ways of configuring custom logging outputs, and they document a way of automating log rotation and compression from within the codebase itself: https://github.com/uber-go/zap/blob/master/FAQ.md#usage.
The text was updated successfully, but these errors were encountered:
A simple solution would be to add optional logging driver config to cdvn repo. Either manually to each compose service or globally as docker default logging driver. A good option is journald which has log rotation built in.
Write debug logs on disk if `--log-output-path` is specified.
Log files are rotated automatically once the file size reaches 50MB.
Rotated log files will be kept for 7 days - max 10 rotated files will be kept, whatever comes first.
category: feature
ticket: #2671
Write debug logs on disk if `--log-output-path` is specified.
Log files are rotated automatically once the file size reaches 50MB.
Rotated log files will be kept for 7 days - max 10 rotated files will be kept, whatever comes first.
category: feature
ticket: #2671
🎯 Problem to be solved
Charon has been designed from the beginning with a no-disk-write approach: aside for the DKG and
create cluster
, Charon never writes anything on disk.The container-exclusive deployment strategy leaves the burden of log retention to the user: since on container restart logs are lost, the user needs to devise a plan to persist them somewhere in order to audit and eventually find and report bugs to us.
🛠️ Proposed solution
I propose we consider the implementation of a on-disk log retention strategy, where Charon writes logs on the already existing facilities (Loki, stdout) as well as files - plural here because log files need to be periodically rotated.
Zap provides multiple ways of configuring custom logging outputs, and they document a way of automating log rotation and compression from within the codebase itself: https://github.com/uber-go/zap/blob/master/FAQ.md#usage.
The text was updated successfully, but these errors were encountered: