Teleport in Amphere K8s get CrashLoopBackOff

gravitational / teleport

The easiest, and most secure way to access and protect all of your infrastructure.

https://goteleport.com

GNU Affero General Public License v3.0

17.42k stars 1.74k forks source link

Teleport in Amphere K8s get CrashLoopBackOff #9348

Closed fdelucchijr closed 1 year ago

fdelucchijr commented 2 years ago

Description

What happened: When I try to run Teleport on a K8s node at Oracle in an Amphere A1 2vcpu 12 RAM it gets the following error: /usr/local/bin/teleport: error while loading shared libraries: libpthread.so.0: ELF load command alignment not page-aligned This is probably related to this.

What you expected to happen: Simply start teleport to manage my k8s internal apps 😢.

Reproduction Steps

As minimally and precisely as possible, describe step-by-step how to reproduce the problem.

Install terraform-cluster helm chart in any amphere cluster

Server Details

Teleport version: is running quay.io/gravitational/teleport:8.0.3 image
Server OS: Ubuntu 20.04, but this shouldn't be a problem because containerization
Where are you running Teleport?: Oracle Amphere OKE Cluster

webvictim commented 2 years ago

All of the Teleport images hosted on https://quay.io/repository/gravitational/teleport?tab=tags (including quay.io/gravitational/teleport:8.0.3) contain amd64 binaries. It appears that the A1 is an ARM-based CPU, meaning that amd64 binaries wouldn't run on it.

You'd need to build a container with Teleport arm or arm64 binaries instead (available for download here) and change the chart to use this image instead.

Related to #3384

zmb3 commented 1 year ago

We now offer ARM container builds.

fdelucchijr commented 1 year ago

Thanks!!!