Problem installing Rapyuta at the "provision all"-step.

[Pieterjan van Gastel]

Hello everyone in the Rapyuta Google Group!

I have a problem installing Rapyuta. I cannot get it to work.

I follow the steps described here: http://rapyuta.org/Install

After installing the packages (which succeeds), I have to setup a container filesystem by doing: $ ./setup/provision all

When it asks for: "Do you want to auto-provision/recreate credentials database for developer mode (Insecure) [y/N]:"

I answer: y

When it asks for: "Enter the root directory to store the Rapyuta Container Filesystem [/opt/rce/container]:"

I answer: rapyutaTest

When it asks for: "Enter the Ubuntu release to deploy in the container."

I answer: 0 (i.e. precise (12.04 LTS))

When it asks for: "Enter the ROS release to deploy in the container."

I answer: 0 (i.e. fuerte)

When it asks for: "Enter the deployment platform for easier auto configuration."

I answer: 0 (i.e. aws)

Then it does multiple Hit, Get, Ign and other stuff.

Finally, after all that, it prints the following:

The following packages were automatically installed and are no longer required:
  ipmitool ros-fuerte-geometry-experimental ros-fuerte-wge100-driver ros-fuerte-openni-camera gir1.2-ubuntuoneui-3.0 ros-fuerte-octomap-msgs nvidia-cg-toolkit
  ros-fuerte-bond-core libgstreamer-plugins-base0.10-dev ros-fuerte-nodelet-core libv8-3.7.12.22 libfltk-cairo1.3 ros-fuerte-sql-database libestools2.1 gforth
  libpcre++-dev ros-fuerte-wifi-drivers ros-fuerte-image-common daemontools libxxf86vm-dev ros-fuerte-flann libgstreamer0.10-dev
  ros-fuerte-visualization-common libfltk-gl1.3 festlex-cmu tcsh ros-fuerte-client-rosjava-jni ros-fuerte-octomap sysstat festvox-kallpc16k
  ros-fuerte-pr2-ethercat-drivers ros-fuerte-camera1394 x11proto-xf86vidmode-dev openipmi python-mechanize libpcre++0 ros-fuerte-ros-realtime
  ros-fuerte-common-tutorials ros-fuerte-ompl ros-fuerte-imu-drivers ros-fuerte-geometry-tutorials ros-fuerte-robot-model-visualization libv8-dev libopenipmi0
  libjpeg62 libubuntuoneui-3.0-1 ros-fuerte-bfl ros-fuerte-audio-common gforth-lib festival ros-fuerte-protobuf ros-fuerte-robot-model-tutorials
  ros-fuerte-prosilica-driver scons festlex-poslex gforth-common ros-fuerte-openni-tracker ros-fuerte-executive-smach-visualization ros-fuerte-camera-drivers
  libopenni-nite-dev ros-fuerte-wg-common libcg ros-fuerte-physics-ode ros-fuerte-common ros-fuerte-multimaster-experimental ros-fuerte-geometry-visualization
Use 'apt-get autoremove' to remove them.
0 upgraded, 0 newly installed, 0 to remove and 3 not upgraded.
 * ovsdb-server is already running
 * ovs-vswitchd is already running
 * iptables already has a rule for gre, not explicitly enabling
debootstrap is /usr/sbin/debootstrap
Checking cache download in /var/cache/lxc/precise/rootfs-amd64 ...
Copy /var/cache/lxc/precise/rootfs-amd64 to rapyutaTest/rootfs ...
Copying rootfs to rapyutaTest/rootfs ...
Please change the root password! (currently, the password for root is root)
Provisioning other required directories and files
Could not setup the rce environment.
There seems to be something wrong with AWS or configuration settings. Please check your configuration.

I have tried to change the root password by doing "sudo passwd root", but it still says "Please change the root password! (currently, the password for root is root)"

I have tried other deployment platforms, and they do not work either.

I have no idea how to check the configuration. I cannot find any troubleshooting information about it.

Any help is appreciated! Thanks in advance!

Best, PJ

[Pieterjan van Gastel]

Here is some extra information.

On http://rapyuta.org/Install it says "For testing purposes, press y to the first prompt and press Enter for the rest."

If I do this, I get the following output: (last lines only)

The following packages were automatically installed and are no longer required:

  ipmitool ros-fuerte-geometry-experimental ros-fuerte-wge100-driver ros-fuerte-openni-camera gir1.2-ubuntuoneui-3.0 ros-fuerte-octomap-msgs nvidia-cg-toolkit

  ros-fuerte-bond-core libgstreamer-plugins-base0.10-dev ros-fuerte-nodelet-core libv8-3.7.12.22 libfltk-cairo1.3 ros-fuerte-sql-database libestools2.1 gforth

  libpcre++-dev ros-fuerte-wifi-drivers ros-fuerte-image-common daemontools libxxf86vm-dev ros-fuerte-flann libgstreamer0.10-dev

  ros-fuerte-visualization-common libfltk-gl1.3 festlex-cmu tcsh ros-fuerte-client-rosjava-jni ros-fuerte-octomap sysstat festvox-kallpc16k

  ros-fuerte-pr2-ethercat-drivers ros-fuerte-camera1394 x11proto-xf86vidmode-dev openipmi python-mechanize libpcre++0 ros-fuerte-ros-realtime

  ros-fuerte-common-tutorials ros-fuerte-ompl ros-fuerte-imu-drivers ros-fuerte-geometry-tutorials ros-fuerte-robot-model-visualization libv8-dev libopenipmi0

  libjpeg62 libubuntuoneui-3.0-1 ros-fuerte-bfl ros-fuerte-audio-common gforth-lib festival ros-fuerte-protobuf ros-fuerte-robot-model-tutorials

  ros-fuerte-prosilica-driver scons festlex-poslex gforth-common ros-fuerte-openni-tracker ros-fuerte-executive-smach-visualization ros-fuerte-camera-drivers

  libopenni-nite-dev ros-fuerte-wg-common libcg ros-fuerte-physics-ode ros-fuerte-common ros-fuerte-multimaster-experimental ros-fuerte-geometry-visualization

Use 'apt-get autoremove' to remove them.

0 upgraded, 0 newly installed, 0 to remove and 3 not upgraded.

 * ovsdb-server is already running

 * ovs-vswitchd is already running

 * iptables already has a rule for gre, not explicitly enabling

debootstrap is /usr/sbin/debootstrap

Checking cache download in /var/cache/lxc/precise/rootfs-amd64 ... 

Copy /var/cache/lxc/precise/rootfs-amd64 to /opt/rce/container/rootfs ... 

Copying rootfs to /opt/rce/container/rootfs ...

Please change the root password! (currently, the password for root is root)

Provisioning other required directories and files

Could not setup the rce environment.

Either the interface 'eth0' isn't connected or there seems to be a network error Please check your configuration.

Also, why would I need the eth0 interface anyway?

[Dominique Hunziker]

Hi Pieterjan,

It seems to me the problem is in the network configuration. In the first mail you select 'aws' (i.e. for amazon's EC2 instances) and in the second mail you select 'other' and then the default settings (i.e. eth0) for a network configuration common for a desktop PC with a wired connection.
Could you specify where you want to install Rapyuta and what the output of ifconfig is?

eth0 is usually the network interface associated with the (external) wired connection. Rapyuta therefore tries to use this network interface for connections to robots. The used network interface then needs a valid IP address, otherwise Rapyuta won't be able to use the network interface, which might also be a reason of the configuration problem.
Also to do some testing you don't necessarily have to change the root password (this concerns the root password of the container; if you changed the root password of the host machine you definitely have to disable/change that).

Dominique

[Pieterjan van Gastel]

Hello Dominique,

I want to install Rapyuta on my laptop to test how it works.

Eventually I want to use it to set up a RoboEarth server which runs the Rapyuta Cloud Engine (probably on a desktop PC, but not sure yet, maybe test it first on my laptop too, I'd have to talk to my supervisor about that..), and then connect one or two ROS-based robots to it.

The output of ifconfig is:

eth0      Link encap:Ethernet  HWaddr e8:03:9a:01:6a:f3  

          UP BROADCAST MULTICAST  MTU:1500  Metric:1

          RX packets:0 errors:0 dropped:0 overruns:0 frame:0

          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:1000 

          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)

lo        Link encap:Local Loopback  

          inet addr:127.0.0.1  Mask:255.0.0.0

          inet6 addr: ::1/128 Scope:Host

          UP LOOPBACK RUNNING  MTU:16436  Metric:1

          RX packets:11206 errors:0 dropped:0 overruns:0 frame:0

          TX packets:11206 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:0 

          RX bytes:1157685 (1.1 MB)  TX bytes:1157685 (1.1 MB)

lxcbr0    Link encap:Ethernet  HWaddr 8a:b6:0c:50:15:b8  

          inet addr:10.0.3.1  Bcast:10.0.3.255  Mask:255.255.255.0

          inet6 addr: fe80::88b6:cff:fe50:15b8/64 Scope:Link

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:0 errors:0 dropped:0 overruns:0 frame:0

          TX packets:71 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:0 

          RX bytes:0 (0.0 B)  TX bytes:10814 (10.8 KB)

wlan0     Link encap:Ethernet  HWaddr dc:a9:71:b8:39:1d  

          inet addr:192.168.1.6  Bcast:192.168.1.255  Mask:255.255.255.0

          inet6 addr: fe80::dea9:71ff:feb8:391d/64 Scope:Link

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:265953 errors:0 dropped:0 overruns:0 frame:0

          TX packets:209432 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:1000 

          RX bytes:305575401 (305.5 MB)  TX bytes:35092695 (35.0 MB)

Does what you're telling me mean that I can use any network interface to connect with the robots, or do I always have to use an ethernet connection?

And do I really need to connect to robots to test how Rapyuta works? Or can I also test it with only my laptop?

Can I also use Rapyuta with simulated robots, such as in the ROS Gazebo simulator?

Best,

Pieterjan

[Dominique Hunziker]

So the problem is as stated that the eth0 interface is not connected/has no IP address. Just use wlan0 instead of eth0 for external & internal communication and you should be fine to run some tests.

The master process of Rapyuta assigns each robot a robot process (responsible for communication with robots) on initial connection attempt. This assignment happens by sending the IP address of the robot process to the robot. The IP address is retrieved based on the definition during setup or at a later time in the config file of the "external Interface" (as far as I know it should even be possible to hardcode an IP address for this instead of specifying an interface; except of the loopback address/interface which should break the code). So it does not have to be an ethernet connection; you could also use bridge/virtual interface, etc. (This is necessary for eg if more than one machine is used to run Rapyuta, wirewall settings, etc.)
The definition of a robot in Rapyuta's sense is just that a robot is a client on the "outside". Therefore it is of course possible to run "the robot" on the same machine as Rapyuta.
So to come back to the interface issue: the external interface is just the interface where Rapyuta can be reached by robots, the internal interface is the interface used for Rapyuta internal communication and the container interface is the network interface used for the lxc's which in ubuntu defaults to lxcbr0.

[Pieterjan van Gastel]

Thanks for your help.

[Spring Mac]

Dear Gastel,

I have a similar problem during rapyuta installation. I have some errors and fails after running " ./setup/provision all". When I tested my installation by running "rce master" first, there was an error occurring, saying, 

"The network device 'lxcbr0' does not exist on your system check your configuration."  And now my network connection is eth0, how can I change eth0 to lxcbr0 ?

Thanks a lot.

Regards,

Spring

Pieterjan van Gastel於 2013年11月13日星期三 UTC+8下午11時03分46秒寫道：

[Dominique Hunziker]

Dear Spring,

The network device 'lxcbr0' is created when the deb package 'lxc' is installed and should be available in any case. If you changed the default configuration for the lxc bridge you will have to adapt accordingly;
otherwise it seems the provision step failed.

Regardless, the network configuration can be changed by editing ~/.rce/config.ini.

Regards,
Dominique

[Spring Mac]

Dear Dominique,

Thank you very much. Could I ask you another question? 

The container file system can't be installed. When I ran "./setup/provision all", some errors happened as following:

Err http://ppa.launchpad.net precise/main Sources                              

  404  Not Found

Err http://ppa.launchpad.net precise/main amd64 Packages                       

  404  Not Found

Err http://ppa.launchpad.net precise/main i386 Packages                        

  404  Not Found

W: GPG error: http://extras.ubuntu.com precise Release: The following signatures were invalid: BADSIG 16126D3A3E5C1192 Ubuntu Extras Archive Automatic Signing Key <ftpm...@ubuntu.com>

W: Failed to fetch http://ppa.launchpad.net/arne-alamut/freenect/ubuntu/dists/precise/main/source/Sources  404  Not Found

W: Failed to fetch http://ppa.launchpad.net/arne-alamut/freenect/ubuntu/dists/precise/main/binary-amd64/Packages  404  Not Found

W: Failed to fetch http://ppa.launchpad.net/arne-alamut/freenect/ubuntu/dists/precise/main/binary-i386/Packages  404  Not Found

E: Some index files failed to download. They have been ignored, or old ones used instead.

Traceback (most recent call last):

  File "./setup/provision", line 472, in <module>

    provision_container(config)

  File "./setup/provision", line 321, in provision_container

    check_call(cmd, shell=True)

  File "/usr/lib/python2.7/subprocess.py", line 511, in check_call

    raise CalledProcessError(retcode, cmd)

subprocess.CalledProcessError: Command 'sudo apt-get update && sudo apt-get -y install curl python-setuptools python-dev' returned non-zero exit status 100

During installing container file system, at the step "Enter the root directory to store the Rapyuta Container Filesystem [/opt/rce/container]: ", I tried to answer "/opt/rce/container", "/home/rce/container" respectively. But, 

when I ran "sudo rce-make", I can't enter the container file system. 

Error: Could not setup the rce environment. Container file system does not exist: /home/rce/container/rootfs Please check your configuration.

Regards,

Spring

[Dominique Hunziker]

The container filesystem is never created because the provision is not completed. When you run the provision the script tries to update the package index but fails as the 404 errors indicate as well as the exit status 100. Therefore the provision is aborted. This particular error can happen in two places either in the host system when the script tries to install the lxc package and others (as this was your first error I believe this is the issue) or the in the container filesystem to install the ROS dependencies.
Are you able to run the following without errors?
$ apt-get update
$ apt-get install lxc

Dominique

[Spring Mac]

Thanks. I can't. There are errors.

/$ apt-get update

E: Could not open lock file /var/lib/apt/lists/lock - open (13: Permission denied)

E: Unable to lock directory /var/lib/apt/lists/

E: Could not open lock file /var/lib/dpkg/lock - open (13: Permission denied)

E: Unable to lock the administration directory (/var/lib/dpkg/), are you root?

/$ apt-get install lxc

E: Could not open lock file /var/lib/dpkg/lock - open (13: Permission denied)

E: Unable to lock the administration directory (/var/lib/dpkg/), are you root?

Spring

Dominique Hunziker於 2015年3月12日星期四 UTC+8下午9時12分44秒寫道：

[Dominique Hunziker]

You have to run these commands as well as the provision script as a super-user, ie use sudo.

[Spring Mac]

I am sorry. There were errors when I ran $ sudo apt-get update . And there was no error when I ran  $ sudo apt-get install lxc . But the container still couldn't be installed. The same errors occurred, errors 404....

Dominique Hunziker於 2015年3月12日星期四 UTC+8下午11時01分08秒寫道：

[Dominique Hunziker]

Yes, the provision is aborted in case sudo apt-get update fails. You will have to fix the package list first.

[Spring Mac]

Dear Dominique,

Thank you again. I think I have solved error 404 and GPG. And I installed container file system again. Many files were downloaded and installed, but there was a new problem:

E: Unable to correct problems, you have held broken packages.

sudo: unable to resolve host ubuntu

/bin/bash: line 1: git: command not found

/bin/bash: line 1: rosdep: command not found

Regards, 

Spring

Dominique Hunziker於 2015年3月12日星期四 UTC+8下午11時20分10秒寫道：

[Dominique Hunziker]

You seem to still have errors as the first line (E: ...) is an error message of apt-get. However, this time the errors are in the container filesystem. Are you able to run rce-make now to enter the container file system?

[Spring Mac]

No, I can't.  error:

bash: /opt/ros/hydro/setup.sh: No such file or directory

However, actually the setup.sh exists in hydro.

Spring

Dominique Hunziker於 2015年3月16日星期一 UTC+8下午4時49分01秒寫道：

[Dominique Hunziker]

I assume the problem is not the host filesystem, but the container filesystem which is (in a default installation) located at /opt/rce/container/rootfs. And in this filesystem ROS has not been installed as the error message seems to indicate.

[Spring Mac]

I want to uninstall rce and then reinstall it. I tried $ sudo apt-get remove rce, but it didn't work. What command should I used?

Dominique Hunziker於 2015年3月16日星期一 UTC+8下午5時09分58秒寫道：

[Dominique Hunziker]

rce is not a debian package, so apt-get does not work.
You can remove the /opt/rce directory and the ~/.rce directory which will remove all config files, etc.
If you want to remove the Python code you will have to look into the location where setuptools installs the packages ( most likely: /usr/local/lib/python2.7/dist-packages ) and manually remove the directories and the packages from the registery as setuptools does not have an uninstall feature ( have a look here: http://stackoverflow.com/questions/1231688/how-do-i-remove-packages-installed-with-pythons-easy-install )
If you want to remove the system dependencies you will have to look at the provision script to compile a list of installed dependencies.

[Spring Mac]

Thank you very much.

I reinstalled rapyuta, but failed with the same errors. 

Could I ask you another question? How to use the Syntactic Query of RoboEarth Web Interface? What type of content should I input in the box of syntactic query?

Dominique Hunziker於 2015年3月16日星期一 UTC+8下午6時36分46秒寫道：

[Sri Ramya]

Hello, 

I am trying to install Rapyuta on my laptop Ubuntu 14.04.3 LTS. I have ROS Jade version installed on my computer. 

When i type the ./setup/provision all

it gives me earlier versions as options for installing. It is going to be a problem? 

Also I am unable to access ~/.rce/config.ini.  It gives me an error "bash: /home/sriramya/.rce/config.ini: No such file or directory". I suspect the setup provision is not done correctly because of the version differences. How do I sort this out? 

[Sri Ramya]

Traceback (most recent call last):

  File "./setup/provision", line 436, in <module>

    _get_container_ros_release())

  File "./setup/provision", line 200, in _get_compat_ros_release

    host_ros = _SUPPORT_VERSIONS[host_ubuntu][0]

KeyError: 'trusty'

This is the error I get when i do ./setup/provision all. Did anyone face the same problem or know the solution to this? I am trying to install Rapyuta on my laptop Ubuntu 14.04.3 LTS. I have ROS Jade version installed on my computer. 

 

Thanks

[Dhananjay Sathe]

Dear Sri Ramya 

This is the Alpha version of the cloud engine and is hereby deprecated. A lot has changed since we initially started this project during our university, both from the perspective of the needs and requirements of robots and the underlying technologies RoboEarth Cloud Engine relied on. After a detailed analysis of the current scenario, at this point it is impractical to build on the existing code base and have decided to take a clean room approach to build something new. Drawing from our experience and the success and shortcomings of building the first edition and what we learnt in the interim we are currently in the process of building a faster, more scalable, secure and fault tolerant version of the cloud engine. 

We are excited to inform you that our team at Rapyuta-Robotics is hard at work to release an all new version of our work as "Rapyuta Core" in the coming months into the open-source domain. When ready we will update this repository to reflect the same and drop a mail on the list to let you all know.

Thank you for your support

Best

D