Minutes:
- Wrappers (Luke):
- Spoke with the developer of SDC's logging wrapper.
- They've been approached by somebody working for Zahi Kupeluto.
- It seemed the conversation was ongoing, but that the wrapper (which exists to abstract EELF) may continue to exist in some capacity.
- To be continued.
- Issues with upstreaming Filebeat work, as the M4 codefreeze looms. (Roger, Michael).
- Discussed issues with EELF (Luke):
- Non-standard, so idioms are unfamiliar and usage varies a lot.
- Misuse of logger names by special loggers, so that messages are attributed to an AT&T package instead of their actual source.
- Tepid agreement on invocation ID:
- Confident that invocation ID can do what we want.
- Some people still need some convincing that it can't be achieved with existing MDCs.
- Argument comes down to the fact that:
- Semantics of existing MDCs is not clearly defined, so their usage varies.
- Empirically we can't achieve it; we're having to rely on heuristics and special cases.
- Moved forward, but not yet closed.
- EELF changes:
- Suggested adding a Marker of the same name as the logger, e.g. security → SECURITY, audit → AUDIT.
- That allows non-EELF logging to produce logs with the same set of Markers (but without attributing messages to AT&T packages).
- Haven't yet:
- Determined the process for making changes to EELF.
- Found resources the work.
- Luke on leave until the 13th. Michael will cover.
- M4 review pending.
M4 review update:
- Checklist completed
- Actions:
- Followup on the readthedocs section, due from the ONAP team by 25th September.
- JIRA cleanup to close complete issues, move incomplete issues to the backlog.
Actions:
- Readthedocs → Mathew Harffy.
- JIRA cleanup → Borislav.
- Any other M4 followup.