Error "unable to find user user"

[t-moe]

Description of problem: When creating a new project on step3 (Injecting and starting workspace agent) i get the error:

[STDERR] Unable to find user user

Che version: 4.0.0-RC13
Che install directory: ~/eclipse-che-4.0.0-RC13-SNAPSHOT
Docker version:

Client:
Version: 1.10.2
API version: 1.22
Go version: go1.5.3
Git commit: c3959b1
Built: Mon Feb 22 21:30:48 2016
OS/Arch: linux/amd64

Server:
Version: 1.10.2
API version: 1.22
Go version: go1.5.3
Git commit: c3959b1
Built: Mon Feb 22 21:30:48 2016
OS/Arch: linux/amd64

Java version:

java version "1.8.0_74"
Java(TM) SE Runtime Environment (build 1.8.0_74-b02)
Java HotSpot(TM) 64-Bit Server VM (build 25.74-b02, mixed mode)

Environment details (AWS, VirtualBox, physical, etc.):

Kernel Version: 3.2.0-4-amd64
Operating System: Debian GNU/Linux 7 (wheezy)
OSType: linux
Architecture: x86_64
CPUs: 2
Total Memory: 1.965 GiB
The my user foo has the id 1000, and permissions to access docker.

Steps to Reproduce:

1. start che locally ./bin/che.sh as user foo(id 1000)
2. create a new project from stack "node".
3. On step 3, you'll get the error: [STDERR] Unable to find user user

Other way to produce a similar error:

1. start che in a docker container using: ./bin/che.sh -i
2. you will get the output

Starting Che server in docker container named che.
Found a container named che. Attempting restart.
Successful restart of container named che. Restarting Che server...
Unable to find user user

[ghost]

Interesting...

What if you run this image manually:

docker run -ti codenvy/node bash

Will you be able to enter a container as a user?

FYI, I was unable to reproduce it on a fresh Debian 8 instance with the latest Docker installed.

[t-moe]

If I run this manually I get into the shell of the container as expected.

How can I find out which steps are executed in "step3 injecting....". ?

This might be totally unrelated but:
./bin/che.sh -p:4433 -r:<IP> -d
gives me

ERROR: transport error 202: bind failed: Address already in use
ERROR: JDWP Transport dt_socket failed to initialize, TRANSPORT_INIT(510)
JDWP exit error AGENT_ERROR_TRANSPORT_INIT(197): No transports initialized [debugInit.c:750]
FATAL ERROR in native method: JDWP No transports initialized, jvmtiError=AGENT_ERROR_TRANSPORT_INIT(197)
./bin/che.sh: line 461: 29564 Aborted "${ASSEMBLY_BIN_DIR}"/catalina.sh jpda ${CHE_SERVER_ACTION}

Where as starting it without the -d flag, raises no error.

[ghost]

When a workspace agent is injected, we unzip it from /mnt/che into /home/user/che - https://github.com/eclipse/che/blob/master/assembly/assembly-ide-war/src/main/java/org/eclipse/che/api/deploy/ApiModule.java#L125

Address already in use is quite self explaining - there seems to be a process listening on the debug port - 8000

[ghost]

Another text you can do is:

docker run -d --name=test codenvy/node
docker exec test cat /etc/hosts

The exec command can be anything you want. I want to make sure execs are working ok.

Looks like you have got an old Kernel. I suspect this may be related - RocketChat/Rocket.Chat#1530

Can you run uname -a?

[t-moe]

$ docker run -d --name=test codenvy/node
$ docker exec test cat /etc/hosts
Unable to find user user

uname -a
Linux <HOSTNAME> 3.2.0-4-amd64 #1 SMP Debian 3.2.73-2+deb7u1 x86_64 GNU/Linux

[ghost]

It is a Kernel issue then. Docs docs say:

Docker requires a 64-bit installation regardless of your Debian version. Additionally, your kernel must be 3.10 at minimum.

[t-moe]

You were right. I upgraded the kernel to 3.16 and now it works.
Thanks a lot!