Trace
In functions which support this level, details every parameter and operation to help diagnose subtle bugs.
This should only be enabled for specific areas of interest or the log volume will be overwhelming.
Some system performance degradation should be expected.
Define TRACE logging level
TRACE is a logging keyword that is understood by most logging tools.
OpenStack has repurposed this in the past to not be TRACE logging
but instead be used whenever a Stacktrace was dumped.
Stack traces should be logged at ERROR level (they currently aren't).
TRACE should be defined as log level 5 in python (which is lower than DEBUG),
and LOG.trace support should be added to oslo logger 1.8.0.
LOG.trace can then be used for deep tracing of code.
Oslo.log logging level
(1) Critical :
ZOMG! Cluster on FIRE! Call all pagers, wake up everyone.
This is an unrecoverable error with a service that has or probably will lead to service death or massive degredation.
(2) Error:
Serious issue with cloud, administrator should be notified immediately via email/pager. On call people expected to respond.
(3) Warning:
Something is not right, should get looked into during the next work week.
Administrators should be working through eliminating warnings as part of normal work.
(4) Info:
normal status messages showing measureable units of positive work passing through under normal functioning of the system.
Should not be so verbose as to overwhelm real signal with noise. Should not be continuous "I'm alive!" messages.
(5) Debug:
developer logging level, only enable if you are interested in reading through a ton of additional information about what is going on.
(6) Trace:
In functions which support this level, details every parameter and operation to help diagnose subtle bugs.
This should only be enabled for specific areas of interest or the log volume will be overwhelming.
Some system performance degradation should be expected.
Link: https://github.com/openstack/openstack-specs/blob/master/specs/log-guidelines.rst