Puppet Agent run on master fails with Error: tar xzf
I upgraded our puppet master from v2015.2.3 to v2015.3.1 (running on Centos 6.7). The upgrade appears to have gone successfully, I do not see any obvious errors after the upgrade process ran. After the upgrade, I noticed that puppet runs on the master node were failing. When I manually run puppet agent –t on the master, I get the following error output:
Error: tar xzf /opt/puppetlabs/server/data/staging/pe_repo-puppet-agent-1.3.5/puppet-agent-el-6-i386.tar.gz --strip=5 returned 2 instead of one of  Error: /Stage[main]/Pe_repo::Platform::El_6_i386/Pe_repo::El[el-6-i386]/Pe_repo::Repo[el-6-i386 2015.3.1]/Pe_staging::Deploy[puppet-agent-el-6-i386.tar.gz]/Pe_staging::Extract[puppet-agent-el-6-i386.tar.gz]/Exec[extract puppet-agent-el-6-i386.tar.gz]/returns: change from notrun to 0 failed: tar xzf /opt/puppetlabs/server/data/staging/pe_repo-puppet-agent-1.3.5/puppet-agent-el-6-i386.tar.gz --strip=5 returned 2 instead of one of 
I am able to get a successful run if I remove the
pe_repo::platform::el_6_i386 class from the PE Master Classification Group. I've also copied the .gz file that puppet is having difficulty with to another location and attempted to run the
tar xzf command against it. When I do, I get the following output:
gzip: stdin: unexpected end of file tar: Unexpected EOF in archive tar: Unexpected EOF in archive tar: Error is not recoverable: exiting now
It seems to me that the .gz file is corrupted. Has anyone ran into this before?