Response to kubectl commands are throttled on controller-0 of subcloud
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
StarlingX |
Fix Released
|
High
|
Yuxing |
Bug Description
Brief Description
-----------------
1. It appears that ownership of .kube and its subdirectory is root:root
2. Due to this kubectl is unable to use local-cache and reaches out to api server to find appropriate APIs and groups.
2a. Changed the ownership of .kube and its subdirectories to sysadmin:
3. This causes overload on api-server which invokes throttling of the request to protect itself against the flood of the requests.
Severity
--------
Major
Steps to Reproduce
------------------
1. Log in to sysadmin in subcloud controller
2. Run command: "sysadmin@
Expected Behavior
------------------
Console will return current nodes, service and pods info
Actual Behavior
----------------
Error message shown: "Throttling request took 1.142565451s, request: GET: ..."
Reproducibility
---------------
100% reproducible
System Configuration
-------
Distributed Cloud
Branch/Pull Time/Commit
-------
Latest master
Last Pass
---------
No test availbe
Timestamp/Logs
--------------
2020-06-03
Test Activity
-------------
N/A
Workaround
----------
Update the folder ownership
Changed in starlingx: | |
assignee: | nobody → Yuxing (yuxing) |
Changed in starlingx: | |
importance: | Undecided → High |
tags: | added: stx.4.0 stx.distcloud |
Changed in starlingx: | |
status: | New → Triaged |
Changed in starlingx: | |
status: | Triaged → In Progress |
This is a new issue brings by updating kubenetes from v1.16 to v1.18. Similar issues found in: /github. com/kubernetes/ kubernetes/ issues/ 86462 /github. com/kubernetes/ kubernetes/ issues/ 84169
https:/
https:/
Will try to modify the fold ownership to avoid that error