[Release] logzio APM Collector 1.0.0 #561
Conversation
8naama
force-pushed
the
release/logzio-apm-collector-1.0.0
branch
from
35f9f7e to
6796009
Compare
| level: ${LOG_LEVEL} | ||
|
|
||
| # Exporter from Traces Collector to SPM Collector | ||
| spmForwarderConfig: |
There was a problem hiding this comment.
I believe we need to rethink about the separation of the config spmForwarderConfig from the main traceConfig. It can be confusing when a user will want to add custom configuration.
There was a problem hiding this comment.
I'll think if there is a better way to manage it, perhaps keep the base config as is and only remove\add the exporter from the pipeline
| level: ${LOG_LEVEL} | ||
|
|
||
| # Service Graph configuration | ||
| serviceGraphConfig: |
There was a problem hiding this comment.
Again not sure its the best practice to separate the logic
| memory: 250Mi | ||
|
|
||
| # Configuration for ports | ||
| ports: |
There was a problem hiding this comment.
ports for traces collector? or spm collector?
There was a problem hiding this comment.
I used it for both, but I can split them to be more precise and reduce the amount of ports for SPM since we need less open ports there 👍🏼
| {{- end }} | ||
|
|
||
| {{- $region := lower .Values.global.logzioRegion }} | ||
| {{- if not (or (eq $region "us") (eq $region "eu") (eq $region "uk") (eq $region "ca") (eq $region "au")) }} |
There was a problem hiding this comment.
Shouldn't we fail the installation if this condition is evaluated?
There was a problem hiding this comment.
I thought of case where there would be a new region, it would be easier to add the code then add a full custom endpoint. Do you think it would be better to fail?
| action: insert | ||
| batch: {} | ||
| service: | ||
| extensions: [health_check, pprof, zpages] |
There was a problem hiding this comment.
Can we try to use https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/extension/storage/filestorage for data persistence? and see how it effects the performance? same goes for the spm collector? maybe a value that can control the storage type
| headers: | ||
| user-agent: "{{ .Chart.Name }}-{{ .Chart.Version }}-helm" | ||
| extensions: | ||
| pprof: |
There was a problem hiding this comment.
Not sure if we still need to use this extension, also zpages
| health_check: | ||
| endpoint: :13133 | ||
| receivers: | ||
| jaeger: |
There was a problem hiding this comment.
I think we can remove this legacy jaeger receiver at this point
Description
As part of the Unified helm chart, this is the new chart for APM
logzio-telemetrylogzio-telemetryit's required for metrics scraping per node which is not needed for traces.Values.global.<secretParam>section. This change provides two key benefits:latencyandcallswhich renamesspan.name>>operation_secondsprefix to Service Graph metrics (which was dropped inv109) following product request to keep the naming of metrics as it wasconnector.spanmetrics.legacyMetricNamesfeature gate to keep old names of SPM metrics (which were changed to match the Service Graph metrics names inv109) following product requestcluster-adminrole permissionsotelLogLevel.tplfiles (_helpersand_config)CHANGELOG.mdto keep the readme more cleanWhat type of PR is this?
(check all applicable)
Added tests?