Daily Lab / POC Activities

Extramural:

Modified the RLS pg_hba.conf permissions for databases lrc1000 and rli1000.

Tested RLS configuration
ncphi:/usr/local/globus-4.0.5/bin # ./globus-rls-admin -p rls://ncphi.phgrid.netping rls://ncphi.phgrid.net: 0 seconds

Updated configuration on lab 1003 to allow transfers between Dallas and Tarrant servers

Tested Java Cog Kit Desktop. The Java Cog Kit Desktop includes: GridFTP Gui, Grid Monitor, Workflow Tool, Grid Shell, Image Viewer, Directory Browser
(www.cogkit.org)

Daily Lab / POC Activities

Extramural:

• Researched GridSQL
• Worked on troubleshooting RLS issue (Trust configuration is the possible root cause)
• Updated RLS Postgresql on lab1002 to include:

Database: lrc1000
t_attribute | table
t_date_attr | table
t_flt_attr | table
t_int_attr | table
t_lfn | table
t_map | table
t_pfn | table
t_rli | table
t_rlipartition | table
t_str_attr | table

Database: rli1000
t_lfn | table
t_lrc | table
t_map | table
t_rli | table
t_rlipartition | table
t_sender | table

Daily Lab / POC Activities

Extramural:

• Installed and configured RLS on lab1002. RLS has been configured to use a Postgresql Database.
  
• Currently working on solving the following RLS issues:

gump:~ # /usr/local/globus-4.0.5/bin/globus-rls-admin -p rls://gump.somedomain.net
connect(rls://gump.somedomain.net): globus_rls_client: Globus I/O error: globus_xio: An end of file occurred

gump:~ # /usr/local/globus-4.0.5/bin/globus-rls-admin -p rlsn://gump.somedomain.net
connect(rlsn://gump.somedomain.net): globus_rls_client: IO timeout: globus_io_register_read() timed out after 30 seconds

• Peter is currently investigating methods for extending the MonaLisa pluggable component code.

Daily Lab / POC Activities

• Demonstrated grid functionality to the Project Management Office
• Provided grid documentation to Peter and instructed him on accessing the grid
• Pittsburgh node is still having time sync issues. (Check skew error) Jeremy is working to correct the issue.
• Peter is currently researching extending MonaLisa and WS MDS functionality
  

Demo Commands:

Login:
grid-proxy-init -verify

3rd Party Transfer from Dallas
globus-url-copy -dbg gsiftp://gump.somedomain.net/home/ken/gridtesting gsiftp://tarrant.somedomain.net/tmp/atl-file

Put

globus-url-copy -dbg file:///home/ken/gridtesting gsiftp://tarrant.somedomain.net/tmp/atl-file

RFT of HL7 file from Atlanta to Tarrant

rft -h gump.somedomain.net -f /tmp/rft.xfr

Submit a Job to WSGRAM:
globusrun-ws -submit -c /bin/rm /tmp/hello*
globusrun-ws -submit -c /bin/true
echo $?
globusrun-ws -submit -c /bin/false

WebMDS Services (Show RFT Status)
http://gump.somedomain.net:8080/webmds/

Daily Lab / POC Activities

Extramural:

• Worked on troubleshooting the Dallas RFT issue.
• Assisted Columbia with installing and configuring Globus
• Spoke with Wayne about installing the JHUAPL node
• Worked on Atlanta node RLS configuration.

Daily Lab / POC Activities

Extramural:

• Researched RLS
• Configured RLS on lab 1002 node
• Installed the following on the Dallas server:
  

postgresql-8.1.9-1.2.i586.rpm
postgresql-libs-8.1.9-1.2.i586.rpm
postgresql-server-8.1.9-1.2.i586.rpm
libiodbc-3.52.5-1.i386.rpm
libiodbc-devel-3.52.5-1.i386.rpm

• Configured Globus Container on Dallas node

bubba@gump:~> counter-client -s https://gump.somedomain.net:8443/wsrf/services/CounterService
Got notification with value: 3
Counter has value: 3
Got notification with value: 13
bubba@gump:~>

• Configured RFT on Dallas node. RFT transfers on the Dallas node hangs with no errors, need to troubleshoot the issue.

Daily Lab / POC Activities

Extramural:

Reinstalled the MonaLisa service agents on Dallas, Tarrant, and Atlanta using the standard install version 1.8.6. The VDT installations of MonaLisa have been removed to simplify the installation document.

Assisted Jeremy with installing and configuring MonaLisa on the Pittsburgh grid node. Pittsburgh is currently on-line and showing a status in the MonaLisa map.

Documented the current state of the grid with screen shots.

Daily Lab / POC Activities

Extramural:

• Researched the Replica Location Service
• Assisted Pittsburgh Engineer with Grid Node installation and testing.
  

The Pittsburgh node was able to transfer date from Atlanta and Tarrant. Jeremy is going to verify firewall settings with the Firewall administrators because Atlanta cannot transfer data from the Pittsburgh node. Data transfers are working fine from the Tarrant node.

Daily Lab / POC Activities

Extramural:

• Reinstalled VDT on Lab 1001.

Note: If the installation is run by a non-root user, the certificates will be stored in the /opt/vdt/globus/etc directory instead of the default /etc/grid-security/certificates directory.

If you select "no" during the certificate installation portion of the install, VDT will not install the default trusted certificates. This is OK because the CA setup file shipped by HealthGrid, will create all required directories and install the certificates to the correct locations. If this method is used, the GRID_SECURITY variable located in /opt/vdt/setup.sh needs to be modified to include the correct directory location. (/etc/grid-security) This needs to be done before vdt-control --on is run.

• Modified the Trusted Authority setup of the VDT installation document based on the recent VDT installations.
• Installed and tested AccessGrid with Ken and Tom
  

PHGrid News

Jeremy Espino at the University of Pittsburgh reported that their network security people approved and set up the firewall policies to allow their grid node to interoperate on the PH Research Informatics Grid.

The U. Pitt node comprises the 4th node on this grid.

Daily Lab / POC Activities

Extramural:

Assisted Albert at Columbia University with a VDT grid node installation. He is currently awaiting host and user certificates.

Albert's Questions:
How will the grid be used to transmit Natural Language Processor requests? These requests can be initiated using a job that transfers the language processor data to the correct grid node via GridFTP or RFT.

How will the data be secured? A combination of x509 certificates and native user account credentials. Each certificate will be mapped to a valid user account on the grid node. Access rights can be restricted based on the native operating system's security settings.

Daily Lab / POC Activities

Extramural:

• Tested the GRAM configuration on Lab1002 and Tarrant. The service is not functional on Dallas.
  

bubba@gump:~> globusrun-ws -submit -c /bin/true
Submitting job...Done.
Job ID: uuid:bec03646-f44a-11dc-934a-d19d36a71b58
Termination time: 03/18/2008 17:51 GMT
Current job state: Active
Current job state: CleanUp
Current job state: Done
Destroying job...Done.
bubba@gump:~> echo $?
0

Troubleshooting the Dallas Postgresql install. The package postgresql-libs-8.1.9-1.2 is hanging during the installation.
Met with PHINMS team for an overview of PHINMS

Basic Questions about OGSA-DAI

> 1) If I downloaded the current version of OGSA-DAI, what would I be able to do?

If you download OGSA-DAI you will be able to expose data resources, such as relational or XML databases or files, through one or more OGSA-DAI services. There are also a number of simple data integration scenarios that you can achieve - send the same query to various databases through an OGSA-DAI service. If you download extension pack you will also get some activities that allow you to do joins across relational databases (extension packs allow us to release functionality early that will be integrated into a future OGSA-DAI release). However, with what we currently release you will not be able to satisfy one of your scenarios yet (other than in the simplest way).

> 2) Generally, how difficult would it be?

Exposing a database through an OGSA-DAI service should not be difficult. The documentation we provide is quite intensive. After that it depends on what you want to do.

> 3) Is there any sample code that could be used to do a basic 'hello world' scenario?

With the base distribution you can do better than that. Have a look at the documentation that is on-line, see:

http://www.ogsadai.org.uk/documentation/ogsadai3.0/ogsadai3.0-axis/

If you download the distribution you will get a version of this documentation (though the one on-line is the most up to date).

To download you will have to register (this allows us provide a metrics to our funders). Try it.

+-----------------------------------------------------------------------+
|Mario Antonioletti:EPCC,JCMB,The King's Buildings,Edinburgh EH9 3JZ. |
|Tel:0131 650 5141|mario@epcc.ed.ac.uk|http://www.epcc.ed.ac.uk/~mario/ |
+-----------------------------------------------------------------------+

Daily Lab / POC Activities

Researched the possibility of reusing existing VeriSign certificates within Globus environment

Findings:

• To generate a new VeriSign certificate, use the certificate request program within Globus and email it to VeriSign for signing.
• VeriSign certificates are x.509 compatible
• VeriSign certificates are issued in base 64 ASCII text
• Openssl can be used to convert the certificates from DER to PEM format as needed by Globus
• In order to configure VeriSign as a trusted CA, the intermediate certificate (cert_hash) needs to be downloaded from the Support Resources site and copied to the TRUSTED_CA directory of Globus
• The PKI admin will have access to the Download Resources site
  

VeriSign Contacts:
Bogdan Buta 650-426-3324
VeriSign Tech Support 1-800-579-2848

# Daily Lab / POC Activities

Extramural:

Attempted UberFTP install using VDT:

The UberFTP installation ended up overwriting the /etc/grid-security/certificates directory causing the Globus container to fail. The failure was corrected by reinstalling the Trusted_CA certificates and removing the /etc/grid-security/certificates-34-1 directory.

All entries regarding /opt/vdt/globus, opt/vdt/gpt, and Trusted_CA were removed from the /opt/vdt/setup.sh script in order to avoid future conflicts.

Verified Globus functionality using GridFTP and WS MDS.
The next UberFTP install attempt will be done using the native installation packages.

Daily Lab / POC Activities

Extramural:

• Installed MonALISA on Dallas and Tarrant
• Used RFT to transfer MonALISA install files from lab1002 to Tarrant and Dallas

bubba@gump:~> rft -h some.domain.net -f /tmp/rft.xfr
Number of transfers in this request: 1
Subscribed for overall status
Termination time to set: 60 minutes

Overall status of transfer:
Finished/Active/Failed/Retrying/Pending
0/1/0/0/0

Overall status of transfer:
Finished/Active/Failed/Retrying/Pending
1/0/0/0/0
All Transfers are completed

bubba@gump:~> rft -h some.domain.net -f /tmp/dallas_rft.xfr
Number of transfers in this request: 1
Subscribed for overall status
Termination time to set: 60 minutes

Overall status of transfer:
Finished/Active/Failed/Retrying/Pending
0/1/0/0/0

Overall status of transfer:
Finished/Active/Failed/Retrying/Pending
1/0/0/0/0
All Transfers are completed

• Changed the $INSTALL_LOCATION/MonaLisa/Service/CMD/ml_env FARM_NAME and SHOULD_UPDATE parameters on all nodes.

• Demonstrated grid functionality
• Researching MonALISA framework
• Started researching UGP portal grid software
  

Daily Lab / POC Activities

Extramural:

• Researched Grid monitoring applications

• Installed and evaluated MonALISA on lab1002
• Researched GUI's for GRAM service
  

Daily Lab / POC Activities

Extramural:

• Configured Index Service on Dallas, Tarrant and Atlanta
• Successfully created an RFT transfer job from Atlanta to Tarrant.
  

Note: (WS MDS)8443/TCP port is blocked on Dallas and Tarrant

Grid Monitoring:

• Configured and tested WS MDS on lab1002

Note: The WS MDS User interface is not very impressive

• Researched and downloaded Monalisa for evaluation

• Researched Virtual Organization Resource Selector (VORS) Note: VORS appears to be specific to Open Science Grid nodes.

Daily Lab / POC Activities

Extramural:

• Researched Medicus Software
• Created Globus Applications spreadsheet
• Installed and tested Access Grid 3.1 software

Daily Lab / POC Activities

Extramural:

• Does a GUI exist for GridFTP?

A Windows based GridFTP GUI called, “GridFTP.Net” is currently available. This application was created by the University of Virginia and it can be found at: http://www.cs.virginia.edu/~gsw2c/GridTools. There are no currently working GridFTP GUI’s for Linux. A couple incubator projects have been initiated, but no GA versions have been released.

Current GridFTP GUI Incubator Projects include:

SJTU GridFTP GUI Client (SGGC)

Java CoG Kit Grid FTP GUI

Although it is not a GUI, UberFTP is a text based interactive GridFTP Client. Information on UberFTP usage can be found at: http://www.teragrid.org/userinfo/data/gridftp.php.

• Continued researching Medicus

• Started loading and configuring Medicus prerequisite software on lab 1002

• Install Status:

* GT 4.0.5 (GridFTP, RLS, GSI) - GridFTP Configured, RLS Configured, GSI In Progress
* COG JGlobus - Not Started
* MySQL 4.x and MySQL J/Connector - MySQL 5.0 Installed, MySQL J/Connector 5.1 installed
* MCS only: OGSA/DAI WSRF 2.2 or higher - Not Started
* SSP and MCS only: UnixODBC - Not Started
* Gateway only: PixelMed(included with MEDICUS)- Deployed in /tmp will be moved to /usr/local
* Gateway only: MEDICUS project code - Deployed in /tmp will be moved to /usr/local

Daily Lab / POC Activities

Extramural:

• Prepared grid nodes for the Grid Demonstration meeting

• Demonstrated GRIDFTP file transfers between servers and 3rd party servers

• Demonstrated the security features of Globus

• Researched the Medicus application

Daily Lab / POC Activities

Extramural:

• Worked on configuring RFT on Dallas and Tarrant
• Configured Globus Container on Dallas and Tarrant

Note: Dallas needs Postgresql installed

• Currently troubleshooting Dallas RFT issue

Note: May need to recreate the rftDatabase in order to resolve the issue.
Completed GRAM configuration on lab 1002

caBig Activities

Reported by:

Ashley Jacobs, Software Consultants, Inc.

I last worked on the CDC grid data service generation on Friday.

I ran into some problems, fixed those, then ran into new ones. The first problem I ran into was the unorthodox installation of MySQL. (Files were located in unexpected locations, and the MySQL server, having gone down, couldn’t be restarted.) I fixed that by re-installing MySQL by using RPM packages from the MySQL website made specifically for the RedHat Enterprise Linux version installed on that machine.

Then I worked with caCORE SDK 3.2.1 to generate the data service. I fixed some of the problems I found when I ran the data service generator, but I ran into others.

Today I sent an email (most of it was copied by me below) to the caCORE Workspace Lead to see if he had any insight into the problems.

I could use a WebEx remote-login session today to retrieve the build log to send it to the Workspace Lead.

Daily Lab / POC Activities

• Installed Globus Web Services on lab 1001
  
• Worked on troubleshooting the following Reliable File Transfer error found in the /usr/local/globus-4.0.5/var/container.log on lab 1002:

2008-03-03 14:40:44,895 ERROR service.ReliableFileTransferImpl [main,:69] Unable to setup database driver with pooling.Connection refused. Check that the hostname and port are correct and that the postmaster is accepting TCP/IP connections.
2008-03-03 14:40:45,127 WARN service.ReliableFileTransferHome [main,initialize:97] All RFT requests will fail and all GRAM jobs that require file staging will fail.Connection refused. Check that the hostname and port are correct and that the postmaster is accepting TCP/IP connections.
Starting SOAP server at: https://192.168.0.202:8443/wsrf/services/
With the following services:

[1]: https://192.168.0.202:8443/wsrf/services/AdminService
[2]: https://192.168.0.202:8443/wsrf/services/AuthzCalloutTestService
[3]: https://192.168.0.202:8443/wsrf/services/CASService
[4]: https://192.168.0.202:8443/wsrf/services/ContainerRegistryEntryService
[5]: https://192.168.0.202:8443/wsrf/services/ContainerRegistryService

• Added the following entry to the /var/lib/pgsql/data/pg_hba.conf file :

host rftDatabase "globus" 192.168.0.202/32 md
NOTE: IP Adress changed for privacy reasons.

This allowed the rft command to be run, but it produced a new error:

bubba@gump:/usr/local/globus-4.0.5/share/globus_wsrf_rft_test> rft -h gump.somedomain.net -f /tmp/rft.xfr
Number of transfers in this request: 1
Subscribed for overall status
Termination time to set: 60 minutes

Overall status of transfer:
Finished/Active/Failed/Retrying/Pending
0/0/1/0/0
Error:Error authenticating user at source/dest hostAuthentication failed [Caused by: Operation unauthorized (Mechanism level: Authorization failed. Expected "/CN=host/localhost" target but received "/O=SomeGrid/OU=Globus Toolkit/OU=Globus Names/OU=us.somegrid.us/CN=host/bubba.somedomain.net")] [Caused by: Authentication failed [Caused by: Operation unauthorized (Mechanism level: Authorization failed. Expected "/CN=host/localhost" target but received "/O=SomeGrid/OU=Globus Toolkit/OU=Globus Names/OU=us.somegrid.us/CN=host/bubba.somedomain.net")]]
All Transfers failed !

• The issue was resolved by updating the /tmp/rft.xfr and replacing the localhost entry with a fully qualified domain name. An example of the file can be found below:

####################
true
16000
16000
false
1
true
1
null
null
false
10
gsiftp://atlanta.somedomain.net:2811/home/bubba/gridtesting
gsiftp://tarrant.anotherdomain.net:2811/tmp/rftTest_Done.tmp
####################################

• Below are the results after the change was made:

bubba@gump:/tmp> rft -h gump.somedomain.net -f /tmp/rft.xfr
Number of transfers in this request: 1
Subscribed for overall status
Termination time to set: 60 minutes

Overall status of transfer:
Finished/Active/Failed/Retrying/Pending
0/1/0/0/0

Overall status of transfer:
Finished/Active/Failed/Retrying/Pending
1/0/0/0/0
All Transfers are completed

• Started testing WS GRAM on lab 1002
  
• Current State of WS GRAM test: Additional configuration needed. Test results below

bubba@gump:~> globusrun-ws -submit -c /bin/touch touched_it
Submitting job...Done.
Job ID: uuid:b516d216-e96a-11dc-ac6c-9fcdae28241b
Termination time: 03/04/2008 21:42 GMT
Current job state: Failed
Destroying job...Done.
globusrun-ws: Job failed: Error code: 201
Script stderr:
bubba's password: