azure-container-networking/debug/windows/npm
Hunter Gregory 7e90960ed0
test(scale): [NPM] fix flakes in kwok and capture kernel state on failure (#2249)
* test(kwok): try standard tier for cluster

* Revert "test(kwok): try standard tier for cluster"

This reverts commit f76e50a559.

* test: run kwok as pod

* fix: add execute permission to sh files

* fix: allow scheduling on linux for kwok pod

* fix: wait timeouts and add retry logic

* fix: make sure to reapply kwok nodes if wait fails

* test: print out cluster state if wait fails

* test: prevent kwok from scheduling on windows node

* test: first wait for kwok pods (20 minutes)

* style: rearrange wait check

* fix: scale up kwok controller for reliability

* fix: typo in scaling kwok pods

* fix: check kwok pods running in test-connectivity instead of test-scale

* fix: wait for pods before adding NetPol

* fix: 7 second timeout for windows agnhost connect

* feat: get cluster state on failure

* debug: fake a failure to verify log capture

* fix: bugs in getting cluster state

* fix: remove newline instead of "n"

* Revert "debug: fake a failure to verify log capture"

This reverts commit 24ec927425.

* feat(win-debug): get prom metrics

* fix: leave timeout=5s for win

* style: remove new, unused --connect-timeout parameter

* style: comment

* feat: top node/pod
2023-11-29 15:37:46 -08:00
..
README.md tool: [WIN-NPM] minor updates to windows script, and new linux script (#1685) 2022-11-04 10:52:48 -07:00
pod_exec.ps1 tool: [WIN-NPM] minor updates to windows script, and new linux script (#1685) 2022-11-04 10:52:48 -07:00
win-debug.ps1 tool: [WIN-NPM] minor updates to windows script, and new linux script (#1685) 2022-11-04 10:52:48 -07:00
win-debug.sh test(scale): [NPM] fix flakes in kwok and capture kernel state on failure (#2249) 2023-11-29 15:37:46 -08:00

README.md

Introduction

This script will collect Windows NPM logs and the HNS and VFP state of the cluster and write them to a new local folder.

How to collect logs

In a PowerShell terminal, navigate to the azure-container-networking/debug/windows/npm folder. Make sure your kubectl is configured to point to the cluster you want to collect logs from (az aks get-credentials -g <resource-group> -n <cluster-name>)

Windows

Run .\win-debug.ps1. The script will create a new folder called logs_DATE containing the results.

Linux

Run .\win-debug.sh. The script will create a new folder called logs_DATE containing the results.

Note: You may not be able to unzip logs.zip in Linux since it was compressed in Windows.