Failed to run install_fedora_deps when starting up

I am running my local kubernetes v1.1.4 cluster on Ubuntu 14.04 using vagrant 1.8.1 and virtualbox 4.3.28. Everything worked fine but since last friday 22.01.2016 i keep getting the same error when executing ./cluster/kube-up.sh

==> master:  *  INFO:  sh -- Version 2015.11.09  
==> master:  *  INFO: System Information:  
==> master:  *  INFO:   CPU:          GenuineIntel  
==> master:  *  INFO:   CPU Arch:     x86_64  
==> master:  *  INFO:   OS Name:      Linux  
==> master:  *  INFO:   OS Version:   3.17.4-301.fc21.x86_64  
==> master:  *  INFO:   Distribution: Fedora 21  
==> master:  *  INFO: Installing master  
==> master:  *  INFO: Found function install_fedora_deps  
==> master:  *  INFO: Found function install_fedora_stable  
==> master:  *  INFO: Found function install_fedora_stable_post  
==> master:  *  INFO: Found function install_fedora_restart_daemons  
==> master:  *  INFO: Found function daemons_running  
==> master:  *  INFO: Found function install_fedora_check_services  
==> master:  *  INFO: Running install_fedora_deps()  
==> master: which: no dnf in (/usr/local/sbin:/sbin:/bin:/usr/sbin:/usr/bin:/root/bin)  
==> master:  *  INFO: Adding SaltStack's COPR repository  
==> master:   
==> master:   
==> master: File contains no section headers.  
==> master: file: file:///etc/yum.repos.d/saltstack-salt-fedora-21.repo, line: 1  
==> master: '<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">\n'  
==> master:  * ERROR: Failed to run install_fedora_deps()!!!  
The SSH command responded with a non-zero exit status. Vagrant
assumes that this means the command failed. The output for this command
should be in the log above. Please read the output to determine what
went wrong.

i ssh into master and this is the content of /etc/yum.repos.d/saltstack-salt-fedora-21.repo

<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>302 Found</title>
</head><body>
<h1>Found</h1>
<p>The document has moved <a href="https://copr.fedorainfracloud.org/coprs/saltstack/salt/repo/fedora-21/saltstack-salt-fedora-21.repo">here</a>.</p>
<hr>
<address>Apache/2.4.6 (Red Hat Enterprise Linux) Server at copr.fedoraproject.org Port 80</address>
</body></html>

I get the same error when running the latest v1.2.0-alpha.6. Is this an issue with the salt-fedora21 repo? Can anyone replicate the error?

Edit: related issue can be found here https://github.com/kubernetes/kubernetes/issues/20088

回答1:

Here's a temporary fix:

As pointed out in https://github.com/kubernetes/kubernetes/issues/20088#issuecomment-174930620, the root issue is a change in URL for an underlying dependency reference by salt-bootstrap.

This is fixed in https://github.com/saltstack/salt-bootstrap/pull/738, but the fix has not made it to salt-bootstrap stable.

We can use the develop version:

cd cluster/vagrant
sed -i "s/https:\/\/bootstrap.saltstack.com/https:\/\/bootstrap.saltstack.com\/develop/" provision-minion.sh 
sed -i "s/https:\/\/bootstrap.saltstack.com/https:\/\/bootstrap.saltstack.com\/develop/" provision-master.sh

Tested on Kubernetes 1.1.4.

回答2:

For macos with it's borky bsd sed:

cd cluster/vagrant
sed -e "s/https:\/\/bootstrap.saltstack.com/https:\/\/bootstrap.saltstack.com\/develop/" provision-minion.sh > provision-minion.sh.back
mv provision-minion.sh.back provision-minion.sh
sed -e "s/https:\/\/bootstrap.saltstack.com/https:\/\/bootstrap.saltstack.com\/develop/" provision-master.sh > provision-master.sh.back
mv provision-master.sh.back provision-master.sh