Vagrant install - tensorflow not installing

#1

Hello,

i’ve tried to install a new VM on 2 different physical servers, both servers without firewall,activated using Centos 7.0 and both hopsworks version (1.2 & 1.3) stuck to the same point

==> default: END_OF_FILE
==> default: sudo chef-solo -c /home/vagrant/.karamel/install/solo.rb -j /home/vagrant/.karamel/install/tensorflow__install.json 2>&1 | tee tensorflow__install.log
==> default: echo ‘https://github.com/logicalclocks/tensorflow-chef/tree/1.2/tensorflow::install’ >> succeed_list
==> default: ’ > tensorflow__install.sh ; chmod +x tensorflow__install.sh ; ./tensorflow__install.sh
==> default: ', DAG is stuck here :frowning:
==> default: INFO [2020-07-19 03:58:33,901] se.kth.karamel.backend.machines.MachinesMonitor: Sending pause signal to all machines

i saw this on other comments but was related to bare-metal installation. where mentioned is a problem with github but i can reach the locations (https://github.com/logicalclocks/tensorflow-chef/tree/1.2/tensorflow & https://github.com/logicalclocks/tensorflow-chef/tree/1.3/tensorflow ) without any problem

#2

i checked the install 6 hrs later and this is the end of the install

==> default: INFO  [2020-07-19 03:58:33,901] se.kth.karamel.backend.machines.MachinesMonitor: Sending pause signal to all machines
==> default: INFO  [2020-07-19 19:20:33,648] se.kth.karamel.webservice.KaramelServiceApplication: Bye! Cleaning up first....
==> default: INFO  [2020-07-19 19:20:33,657] org.eclipse.jetty.server.ServerConnector: Stopped karamel-core@19ae6bb{HTTP/1.1}{0.0.0.0:9090}
==> default: INFO  [2020-07-19 19:20:33,659] org.eclipse.jetty.server.handler.ContextHandler: Stopped i.d.j.MutableServletContextHandler@5cad8b7d{/admin,null,UNAVAILABLE}
==> default: INFO  [2020-07-19 19:20:33,659] org.eclipse.jetty.server.handler.ContextHandler: Stopped i.d.j.MutableServletContextHandler@41de5768{/,null,UNAVAILABLE}
==> default: STDERR:
==> default: ---- End output of "bash"  "/tmp/chef-script20200719-13136-cwn5p" ----
==> default: Ran "bash"  "/tmp/chef-script20200719-13136-cwn5p" returned
==> default: [2020-07-19T19:20:36+00:00] FATAL: Chef::Exceptions::ChildConvergeError: Chef run process exited unsuccessfully (exit code 1)
Chef never successfully completed! Any errors should be visible in the
output above. Please fix your recipes so that they properly complete.
#3

Hi @Fernando_Marines,

You can check (inside the vm) under /home/[install user]/.karamel/install the tensorflow__install.log for a more detailed reason for the failure. If you paste the last 20-30 lines of it here, we might be able to help you further.

#4

Thank you @Alex for your reply

reviewing the log i was able to understand what i did wrong,turns out the GPU’s are missing on the Vm and the Nvidia tests fails

^[[0m platform=centos
^[[0m platform_version=7.8.2003
^[[0m ruby=ruby 2.5.3p105 (2018-10-18 revision 65156) [x86_64-linux]
^[[0m program_name=/bin/chef-solo
^[[0m executable=/opt/chefdk/bin/chef-solo^[[0m

^[[0m^[[0m
Running handlers:^[[0m
[2020-07-20T06:32:41+00:00] ERROR: Running exception handlers
Running handlers complete
^[[0m[2020-07-20T06:32:41+00:00] ERROR: Exception handlers complete
Chef Client failed. 1 resources updated in 13 seconds^[[0m
[2020-07-20T06:32:41+00:00] FATAL: Stacktrace dumped to /tmp/chef-solo/chef-stacktrace.out
[2020-07-20T06:32:41+00:00] FATAL: Please provide the contents of the stacktrace.out file if you file a bug report
[2020-07-20T06:32:41+00:00] FATAL: Mixlib::ShellOut::ShellCommandFailed: bash[test_nvidia] (tensorflow::install line 216) had an error: Mixlib::ShellOut::ShellCommandFailed: Expected process to exit with [0], but received '1’
---- Begin output of “bash” “/tmp/chef-script20200720-30809-l2w8gh” ----
STDOUT:
STDERR:
---- End output of “bash” “/tmp/chef-script20200720-30809-l2w8gh” ----
Ran “bash” “/tmp/chef-script20200720-30809-l2w8gh” returned 1