Friday, May 8, 2009

Fighting with NIC and GridFTP account

1. Trying to install it again.



2. Installing autocontainer:

magg@gobo:~/nimbus$ wget http://workspace.globus.org/downloads/autocontainer-GT4.0.8-v1.0.tar.gz; tar xfvz autocontainer-GT4.0.8-v1.0.tar.gz

Following the documentation:

magg@gobo:~/nimbus/autocontainer$ ./bin/setup-container.sh

  • a hostname that clients can use to contact the container: gobo.mathcs.emory.edu
  • in one terminal run:
    $ /home/magg/nimbus/autocontainer/bin/start-container.sh
  • in another terminal run:
    $ source /home/magg/nimbus/autocontainer/bin/source-me.sh
    $ /home/magg/nimbus/autocontainer/gt/ws-core-4.0.8/bin/grid-proxy-init -cert ~/.globus/usercert-1.pem -key ~/.globus/userkey-1.pem
    $ /home/magg/nimbus/autocontainer/bin/test-container.sh
  • Tests passed:
    magg@gobo:~/nimbus$ /home/magg/nimbus/autocontainer/bin/test-container.sh
    Got notification with value: 3
    Counter has value: 3
    Got notification with value: 13
    magg@gobo:~/nimbus$
3. I tested the xen. It looks that the bridge

gobo:/home/xen/domains# xm create hello-cloud/hello-cloud.cfg
Using config file "./hello-cloud/hello-cloud.cfg".
Started domain hellocloud
gobo:/home/xen/domains# brctl show
bridge name bridge id STP enabled interfaces
eth0 8000.000bdb6fe75f no peth0 vif1.0


4. wget http://workspace.globus.org/downloads/nimbus-TP2.2.tar.gz ; tar xfvz nimbus-TP2.2.tar.gz

4a. Setting GLOBUS_LOCATION:

magg@gobo:~/nimbus/nimbus-TP2.2$ source ~/nimbus/autocontainer/bin/source-me.sh

4b. magg@gobo:~/nimbus/nimbus-TP2.2$ ./bin/all-build-and-install.sh

BUILD SUCCESSFUL
Total time: 1 minute 5 seconds
Successful: Nimbus - build and install all

4c. Autoconfiguration:

magg@gobo:~/nimbus/nimbus-TP2.2$ $GLOBUS_LOCATION/share/nimbus-autoconfig/autoconfig.sh

VMM - hypevisor - Pick a VMM to test with, enter a hostname: gobo.mathcs.emory.edu

How much RAM (MB) should be allocated for a test VM on the 'gobo.mathcs.emory.edu' VMM?
128

container account is (magg)

Now we'll set up the *hostname* that VMMs will use to contact the container over SSHd:
gobo.mathcs.emory.edu

We need to interrupt this part of the installation since now we need to install workspace control.

5. Setting workspace control:
Instructions:

magg/magg (privileged user and its group)

gobo:/home/magg/nimbus# ls
autocontainer nimbus-controls-TP2.2.tar.gz old-nimbus
autocontainer-GT4.0.8-v1.0.tar.gz nimbus-TP2.2
nimbus-controls-TP2.2 nimbus-TP2.2.tar.gz
gobo:/home/magg/nimbus# rm -rf nimbus-controls-TP2.2
gobo:/home/magg/nimbus# tar xzf nimbus-controls-TP2.2.tar.gz
gobo:/home/magg/nimbus# cd nimbus-controls-TP2.2
gobo:/home/magg/nimbus/nimbus-controls-TP2.2# mkdir /opt/workspace
gobo:/home/magg/nimbus/nimbus-controls-TP2.2# cp workspace-control/worksp.conf.example /opt/workspace/worksp.conf

5a. First dry run: (option -o)
gobo:/home/magg/nimbus/nimbus-controls-TP2.2# cd workspace-control/

gobo:/home/magg/nimbus/nimbus-controls-TP2.2# python install.py -c /opt/workspace/worksp.conf -a magg -g magg -o

gobo:/home/magg/nimbus/nimbus-controls-TP2.2/workspace-control# python install.py -c /opt/workspace/worksp.conf -a magg -g magg -i

Actually option '-n' also should work even better:
gobo:/home/magg/nimbus/nimbus-controls-TP2.2/workspace-control# python install.py -c /opt/workspace/worksp.conf -a magg -g magg -i

There are some things you have to do after the script manually.
One of them is

Changing lines in or adding new lines to.

/etc/dhcp3/dhcpd.conf

and updating the location of this file in file

/opt/workspace/bin/dhcp-config.sh

5b. Configuring images:

gobo:/home/magg/nimbus/nimbus-controls-TP2.2/workspace-control# cp /home/xen/domains/hello-cloud/hello-cloud.img /opt/workspace/images/

List the image:

gobo:/home/magg/nimbus/nimbus-controls-TP2.2/workspace-control# nano /opt/workspace/worksp.conf

Configuring the networks:

gobo:/home/magg/nimbus/nimbus-controls-TP2.2/workspace-control# nano /opt/workspace/worksp.conf

association_0: public; eth0; eth0 ; none; 170.140.151.0/24
#association_0: public; eth0; eth0 ; none; 192.168.0.0/24
#;association_1: public; eth0; eth0 ; none; 170.140.151.0/23


/etc/dhcp3/dhcpd.conf
subnet 170.140.151.55 netmask 255.255.255.255 {
}

subnet 170.140.151.94 netmask 255.255.255.255 {
}


5c. now testing. First restart dhcp:
/etc/init/dhcp3-server restart

gobo:/home/magg/nimbus/nimbus-controls-TP2.2/workspace-control# wget http://workspace.globus.org/vm/TP2.2/admin/test-create.sh

5d. Change the location of the image:

gobo:/home/magg/nimbus/nimbus-controls-TP2.2/workspace-control# nano /home/magg/nimbus/autocontainer/gt/ws-core-4.0.8/share/nimbus-clients/sample-workspace.xml

file://hello-cloud.img

We are coming back to the configuration

4d. Coming back to the configuration:
Ok Point of no return but: testing for wembley.mathcs.emory.edu

And then manually run '/home/magg/nimbus/autocontainer/gt/ws-core-4.0.8/share/nimbus-autoconfig/autoconfig-adjustments.sh' at your leisure.

There is a note:

NOTE: you need to MATCH this network in the workspace-control configuration file.
This configuration file is at '/opt/workspace/worksp.conf' by default

For example, you might have this line:

association_0: public; xenbr0; vif0.1 ; none; 170.140.151.54/24

... "public" is the name of the network we chose.
... "xenbr0" is the name of the bridge to put VMs in this network on.
... "vif0.1" is the interface where the DHCP server is listening in dom0 on the VMM
... and the network address range serves as a sanity check (you can disable that check in the conf file)

----------

Making sure 'fake mode' is off:

[*] The 'fake.mode' configuration was:
... set to 'false'
... (it used to be set to 'true')
... in the file '/home/magg/nimbus/autocontainer/gt/ws-core-4.0.8/etc/nimbus/workspace-service/other/common.conf'

----------

Finished.

See 'NOTE' above.

4e. Testing

magg@gobo:~/nimbus$ wget http://workspace.globus.org/vm/TP2.2/admin/test-create.sh

4f. I have got the error Unknown CA. Rebooting the machine and testing it once more.

4g. After reboot:
magg@gobo:~/nimbus/autocontainer$ ./bin/setup-container.sh

in one terminal:
magg@gobo:~$ /home/magg/nimbus/autocontainer/bin/start-container.sh

in the other terminal:
magg@gobo:~/nimbus$ source /home/magg/nimbus/autocontainer/bin/source-me.sh
magg@gobo:~/nimbus$ /home/magg/nimbus/autocontainer/gt/ws-core-4.0.8/bin/grid-proxy-init

Now the playing with gridmapfile:

magg@gobo:~/nimbus$ cp /home/magg/.globus/auto-container-01/grid-mapfile /home/magg/nimbus/autocontainer/gt/ws-core-4.0.8/etc/nimbus/nimbus-grid-mapfile

Changing in both files to not_real_account: magg

magg@gobo:~/nimbus/autocontainer$ nano /home/magg/nimbus/autocontainer/gt/ws-core-4.0.8/etc/nimbus/factory-security-config.xml


to


After running the test script I have some problems:

magg@gobo:~/nimbus$ sh test-create.sh

Workspace Factory Service:
https://127.0.0.1:8443/wsrf/services/WorkspaceFactoryService

Read metadata file: "/home/magg/nimbus/autocontainer/gt/ws-core-4.0.8/share/nimbus-clients/sample-workspace.xml"
Created deployment request soley from arguments.

Creating workspace "http://example1/localhost/image"...
Problem: Resource request denied: Error creating workspace(s): No resource pool has an applicable entry
magg@gobo:~/nimbus$

No comments:

Post a Comment