Notes
- was off Fri and part of Mon doctor/dentist
- SPEC lockdown with Lorraine Welch last Thu
- M4 code freeze on Thu 20th
- Full deployment fix being merged for https://lists.onap.org/g/onap-discuss/topic/oom_full_deployment_do_a/25736532?p=,,,20,0,0,0::recentpostdate%2Fsticky,,,20,2,0,25736532
- CLAMP needs to add the "prune" filter to the common chart for logstash - need to coordinate - they are on 6.1.2
- Code to merge
- WIKIs to finish
- JIRAs to finish
LOG-630
-
Getting issue details...
STATUS
- Questions to answer
- DCAE involvement
- Helm deletion example - assist with 2.9.1 workaround
- verify logging repo does not have maven 3.3/3.5 issue Prudence Au found in pomba repos - resolved - thanks Prudence Au
Pending Work
- merge logback.xml
- import portal/sdk library to log war - then add new logback.xml and adjust library to fit
- retrofit slf4j and slf4j-demo to use new logback.xml - specifically each of the 29 log fields
- finish logstash files plugin to remove all filebeat sidecar containers
- move exiting spec + field 29 ( )to dublin spec - ONAP Application Logging Specification - Post Dublin
- Expand (as Acumos has) collection from k8s system and tomcat logs (access logs for example)
- quick script to validate logs fields
- ...
{"serverDuration": 483, "requestCorrelationId": "1208608b23f1a78c"}
2 Comments
Michael O'Brien
Put a JIRA in for this - Q: from Dave Williamson - in the context of is the field "serverFQDN" useful
LOG-666 - Getting issue details... STATUS
TODO: serverFQDN currently posts "localhost" when in a container - we are not getting the container name or the vm name without some code changes - but they are available to the kubernetes system and should be written out by filebeat/logstash
test a failover - especially if logstash "files" is capturing
harvester should know vm name like "a-onap-devops" - so the java library may not need to actually capture the serverFQDN
onap-msb-consul-df5cdcdbb-n8f5g 1/1 Running 0 29m 10.42.171.138 a-onap-devops
verify that the first part of the id changes on a a cluster vm redeploy "df5cdcdbb-n8f5g" - specifically the df5cdcdbb part - as the n8f5g part is the part of the id that "does" vary across pod restarts on a single VM.
all this assumes you are in ELK + logstash "files" plugin or keep "beats" + filebeat - in both cases verify server name is derived from harvester
TODO: Michael O'Brien Post good/bad logs from current spec
https://gerrit.onap.org/r/#/c/62405/13/reference/logging-demo/src/main/resources/logback.xml
TODO: verify which library they use - SO
Michael O'Brien
moving
OOM-1311 - Getting issue details... STATUS
to
LOG-672 - Getting issue details... STATUS