Details
-
Improvement
-
Status: Open
-
Major
-
Resolution: Unresolved
-
None
-
None
-
None
Description
To improve the observability is a key requirement to achieve better correctness and performance with Ozone.
This jira collects some of the tasks which can provide better visibility to the ozone internals.
We have two main tools:
- Distributed tracing (opentracing) can help to detected performance battlenecks
- Ozone insight tool (a simple cli frontend for Hadoop metrics and log4j logging) can help to get better understanding about the current state/behavior of specific components.
Both of them can be improved to make it more powerful.
Attachments
1.
|
Improve the visibility with Ozone Insight tool |
|
Resolved | Marton Elek |
|
|||||||
2.
|
Create generic service facade with tracing/metrics/logging support |
|
Resolved | Marton Elek |
|
|||||||
3.
|
Make StorageContainerDatanodeProtocolService message based |
|
Resolved | Marton Elek |
|
|||||||
4.
|
Create insight point to debug one specific pipeline |
|
Resolved | Marton Elek |
|
|||||||
5.
|
Support filters in ozone insight point |
|
Resolved | Marton Elek |
|
|||||||
6.
|
Make StorageContainerLocationProtocolService message based |
|
Resolved | Marton Elek |
|
|||||||
7.
|
Make SCMSecurityProtocol message based |
|
Resolved | Marton Elek |
|
|||||||
8.
|
Tracing in OzoneManager call is propagated with wrong parent |
|
Resolved | Attila Doroszlai |
|
|||||||
9.
|
Use annotations to define description/filter/required filters of an InsightPoint |
|
Open | Unassigned | ||||||||
10.
|
Support Ozone insight tool in secure cluster |
|
Open | Unassigned | ||||||||
11.
|
Create insight point for datanode container protocol |
|
Resolved | Marton Elek |
