I am using
**systemLog.traceAllExceptions** option for production replicaset. This makes the log daemon to write the complete exception lines including C++ codes. I observed very huge size to my log files due to this exception lines. is it really required to enable this option unless we dig into the log file in higher level. what is the real case we required to have this info?
I am using
The default log level suffices for typical production troubleshooting.
Per the documentation on
systemLog.traceAllExceptions, this option is
false by default:
Default : false
Print verbose information for debugging. Use for additional logging for support-related troubleshooting.
I would only enable this option (or other increased logging) if you are working with a support or development team that has asked for extra debugging detail to investigate a problem, and generally only enable verbose logging long enough to capture a sample of more detailed diagnostic information. If you are encountering the same problem frequently, multiple copies of verbose logging may not provide additional information but will consume a lot of extra disk space.
I also recommend setting up Log Rotation using
logrotate or a similar utility for your O/S so log file sizes remain manageable.
@Stennie appreciate your response… thanks Stennie.