Introduction to Globus with Condor-G

01 May 2003 Globus with Condor-G 1

Introduction to Globus with Condor-G

Itzhak Ben Akiva (TAU)

David Front (WI)

Israel Academic Grid (IAG)


Agenda

• Grid security and certificates

• Globus

• Condor-G

• Condor-G submission examples

• References


Grid Security Infrastructure (GSI)• GSI is a set of tools, libraries and protocols

used in Globus to allow users and applications to securely access resources.

• Based on a public key infrastructure, with certificate authorities and X509 certificates

PKI(CAs and

Certificates)

SSL/TLS

Proxies and Delegation

PKI forcredentials

SSL forAuthenticationAnd message protection

Proxies and delegation for secure singleSign-on


Public Key Infrastructure (PKI)• PKI allows you to know that a given

public key belongs to a given user

• PKI builds off of asymmetric encryption:

– Each entity has two keys: public and private

– Data encrypted with one key can only be decrypted with other.

– The private key is known only to the entity

• The public key is given to the world encapsulated in a X.509 certificate

Owner


Certificates• Certificates link between public key & identity of a:

– Person, organization, or device (“subject”)– Associated with use of private key– Used by a “relying party”

• Certificate Authority (CA) are responsible for establishing identity

• CA generates key pair, and digitally signs the public key making it a Certificate

John Doe755 E. WoodlawnUrbana IL 61801BD 08-06-65Male 6’0” 200lbs

State ofIllinois

Seal

NameIssuerPublic KeySignature


Certificates: what CA to use?The following possibilities go from the less secure to the most secure: 1. Anyone can become a certificate authority:

http://www.onlamp.com/pub/a/onlamp/2003/02/06/linuxhacks.html 2. Free certificate authorities:

– http://www.thawte.com/– http://www.verisign.com/ ...

3. Globus CA:http://www-fp.globus.org/gt2.4/admin/guide-verify.html#cert coerces some security limitations such as:

– Domain of host to get a certificate should be the same as requestor’s email domain4. EDG CA:

http://igc.services.cnrs.fr/Datagrid-fr/english/index.htmlEDG does not honor Globus certificates.EDG Grants certificates only to people that are known personally by an authorized third person.

For each non-Globus CA, each site should configure the CA as being a trusted one.Globus is trusted by default and EDG has RPMs for adding it as a trusted CA.Recommendation: until Machba supplies certificates, use Globus CA.


Globus with Condor-G. A resource broker to be added later


GlobusGurus:

Ian Foster Carl Kesselman

Globus Project™ www.globus.org

Globus is a bag of tools

Grid SW projects use Globus


What is Globus?

• A research and development project that enables the application of Grid concepts to scientific and engineering computing.

• Globus Toolkit allows: build Grids, develop Grid applications • Globus Project research targets technical challenges and

Globus Toolkit supplies a set of services and software libraries to support Grids and Grid applications: – (GRAM) resource management– (GSI) security– (MDS) information infrastructure – (GASS) data management– (HBM) fault detection– (Nexus and globus_io) portability + communication


The Globus Toolkit contents • A “bag of services”:

components to develop grid applications + programming tools• Component have a C application programmer interface (API)• Some Components have Java classes and/or command line tools• Prototypes of

– higher components (resource brokers, co-allocators) – and services.

• Others use the Globus Toolkit to develop: – higher-level services, – application frameworks, – and scientific/engineering applications

Example:– Condor-G uses Globus for its high-throughput computing

framework


Globus Toolkit: GRAM + GSI

• Globus Resource Allocation Manager (GRAM)– Resource allocation – Process creation – Monitoring– Management – Maps requests

expressed in a Resource Specification Language (RSL) into commands

to local schedulers and computers.

• Grid Security Infrastructure (GSI) – A single-sign-on, run-anywhere authentication service, – local control over access rights – mapping from global to local user identities. – Smartcard support increases credential security.


Globus Toolkit: MDS + GASS

• Monitoring and Discovery Service (MDS):– Extensible Grid information service – Combines data discovery mechanisms with the Lightweight

Directory Access Protocol (LDAP). – Uniform framework for providing and accessing system

configuration and status information, such as:• Compute server configuration• Network status, • Locations of replicated datasets

• Global Access to Secondary Storage (GASS):– Implements a variety of automatic and programmer-managed

data movement and data access strategies– Enables remote programs to read and write local data.


Globus Toolkit: Nexus, globus_io, HBM, GPT

• Nexus and globus_io:communication services for heterogeneous environments:– multimethod communication– multithreading– single-sided operations

• The Heartbeat Monitor (HBM):Allows system administrators or ordinary users to– detect failure of system components – detect failure of application processes

• Globus Packaging Tool: (GPT):– VDT’s Packman installs Globus and Condor-G.

Hence, is it more appropriate for Condor-G users.


Layered Grid Architecture(By Analogy to Internet Architecture)

Application

Fabric“Controlling things locally”: Access to, & control of, resources

Connectivity“Talking to things”: communication (Internet protocols) & security

Resource“Sharing single resources”: negotiating access, controlling use

Collective“Coordinating multiple resources”: ubiquitous infrastructure services, app-specific distributed services

InternetTransport

Application

Link

Inte

rnet P

roto

col

Arch

itectu

re


User

Userprocess #1

Proxy

Authenticate & create proxy

credential

GSI(Grid

Security Infrastruc-

ture)

The Globus Toolkit in One SlideGrid protocols (GSI, GRAM, …) enable resource sharing within virtual orgs;

toolkit provides reference implementation

( = Globus Toolkit services)

Protocols (and APIs) are central to Globus toolkit

Other service(e.g. GridFTP)

Other GSI-authenticated remote service

requests

GIIS: GridInformationIndex Server (discovery)

MDS-2(Monitor./Discov. Svc.)

Soft stateregistration;

enquiry

Gatekeeper(factory)

Reliable remote

invocation

GRAM(Grid Resource Allocation & Management)

Reporter(registry +discovery)

Userprocess #2Proxy #2

Create process Register


Globus Toolkit: missing, weak, plans 1

• (GRAM) resource management– Condor-G adds: reliable job submission– (EDG) Resource broker: choose Globus resource to submit job– Globus plans: support end-to-end performance management and

fault tolerance via network scheduling, advance reservations, and policy-based authorization.

• (GSI) securityUsing X.509 certificates has various limitations. For example:– If a user does not use a pass phrase,

anyone that puts a hand on her certificate can use it.• (MDS) information infrastructure

– LDAP: too weak for frequently changing information (EDG uses RDBM instead of LDAP)


Globus Toolkit: missing, weak, plans 2

• (GASS) data management– Globus current replica management capabilities are limited – Globus plans: provide high-performance access

to large amounts of data (terabytes or petabytes). • (HBM) fault detection• (Nexus and globus_io) portability + communication• Others

– Weak accounting – ‘The firewall problem’: In order to submit a Globus job, some Internet

ports should be opened. This is a security problem.– Weak fabric tools:

• Installation/configuration:– VDT adds configurable installation via Packman– EDG adds Client-server installation + updating via LCFG

• Weak monitoring tools– Restricted support for Windows.


Globus support for Windows

• Port of the Globus Toolkit to the Windows XP/2000 platform is under development/test.

• Using Grid resources from Windows systems or turn Windows systems into Grid resources:– The Java CoG Kit (http://www.globus.org/cog/) provides access

to Grid services via the Java programming language, available on Windows.

– A Java-based GRAM service is currently being developed.– The Condor software from the Condor Project at the University

of Wisconsin (http://www.cs.wisc.edu/condor/) provides job management services that allow you to submit jobs to a local service that then submits your jobs to remote resources for execution. Condor can use Grid resources to execute these jobs. Condor is available for Windows.


Globus toolkit (re)structure

GRAM MDS

GSI

GridFTP MDS

GSI

???

GSI

Reliable invocationSoft state

management

Notification

ComputeResource

DataResource

Other Serviceor Application

Jobmanager

Jobmanager

Lots of good mechanisms, but (with the exception of GSI) not that easilyincorporated into other systems

Service naming


Service Oriented Architecture (SOA)

Service Registry

Service Requestor

Service Provider

Bind

PublishFind

New buzzwords:

•Services (in addition to protocols and APIs)

•Open Grid Services Architecture (OGSA)

•Web services

•Soap

•XML

OGSA may become standard


The Grid Service =Interfaces/Behaviors + Service Data

Servicedata

element

Servicedata

element

Servicedata

element

Implementation

GridService(required)Service data query

Explicit destructionSoft-state lifetime

… other interfaces …(optional) Standard:

- Notification- Service creation- Service registry- Authorization- Manageability

+ application-specific interfaces

Binding properties:- Reliable invocation- Authentication

Hosting environment/runtime(“C”, J2EE, .NET, …)


Condor-GGuru:

Miron Livny

Condor Project: http://www.cs.wisc.edu/condor

Condor-G manual: http://www.cs.wisc.edu/condor/manual/v6.4/5_2Condor_G.html

Condor is a scheduler, similar to PBS, LSF and others

Condor-G is (the submission) part of Condor.

It adds to Globus Reliable job submission


Condor-G from Globus eyes• Condor-G adds to Globus reliable job submission. It lets you:

– Submit jobs into a queue– have a log detailing the life cycle of your jobs– manage your input and output files– along with everything else you expect from a job queuing system.

• Condor-G does more than Globus toolkit's globusrun command: – It allows you to submit many jobs at once – and then to monitor those jobs with a convenient interface – receive notification when jobs complete or fail– maintain your Globus credentials which may expire while a job runs

• Condor-G is a fault-tolerant system:If your machine crashes, you can still perform all of these functions when your machine returns to life.


Condor-G from Condor eyes• Condor-G is a Globus-enabled version of the Condor scheduler.

It uses Globus to handle inter-organizational problems like:– Security– Resource management for supercomputers,– Executable staging.

Hence: The same Condor tools that access local resources are now able to use the Globus protocols to access resources at multiple sites.

• Condor-G manages both a queue of jobs and the resources from one or more sites where those jobs can execute. It communicates with these resources and transfers files to and from these resources using Globus mechanisms, such as:

– GSI– GRAM protocol for job submission, – and a local GASS server for file transfers.

• The mutual look:Condor can be used to submit jobs to systems managed by Globus.Globus tools can be used to submit jobs to systems managed by Condor.


how Condor-G interacts with Globus protocols


Submitting a job to Condor-G: example 1

Run your compiled program on a different Globus resource:

• Make sure your Condor server service is running on the Condor server.(Not explained here)

• Make sure you have your Grid credentials, create a proxy: grid-proxy-init

• To submit a job:condor_ submit < submit-description-file-name>

• The following sample runs a job on the Origin2000 at NCSA: executable = test globusscheduler = modi4.ncsa.uiuc.edu/jobmanager universe = globus output = test.out log = test.log queue

• The executable for this example is transferred from the local machine to the remote machine.

• By default, Condor transfers the executable, as well as any files specified by the input command.

• This executable must be compiled for the correct intended platform.


Submitting a job to Condor-G example 1 cont.

• The globusscheduler command is dependent on the scheduling software available on remote resource. This required command will change based on the Grid resource intended for execution of the job.

• All Condor-G jobs are submitted to the globus universe. Hence:universe = globus is always required in the submit description file.

• IO:No input file is specified for this example job. Any output (file specified by the output) or error (file specified by the error) is transferred from the remote machine to the local machine as it is produced. This implies that these files may be incomplete in the case where the executable does not finish running on the remote resource. The job log file is maintained on the local machine.

• To submit this job to Condor-G for execution on the remote machine, use: condor_submit test.submit where test.submit is the name of the submit description file.



Example output from condor_ q for this submission looks like: % condor_q -- Submitter: wireless48.cs.wisc.edu : <128.105.48.148:33012> : wireless48.cs.wi ID OWNER SUBMITTED RUN_TIME ST PRI SIZE CMD 7.0 epaulson 3/26 14:08 0+00:00:00 I 0 0.0 test 1 jobs; 1 idle, 0 running, 0 held

After a short time, Globus accepts the job. Again running condor_ q will now result in

% condor_q -- Submitter: wireless48.cs.wisc.edu : <128.105.48.148:33012> : wireless48.cs.wi ID OWNER SUBMITTED RUN_TIME ST PRI SIZE CMD 7.0 epaulson 3/26 14:08 0+00:01:15 R 0 0.0 test 1 jobs; 0 idle, 1 running, 0 held

Then, very shortly after that, the queue will be empty again, because the job has finished:

% condor_q -- Submitter: wireless48.cs.wisc.edu : <128.105.48.148:33012> : wireless48.cs.wi ID OWNER SUBMITTED RUN_TIME ST PRI SIZE CMD 0 jobs; 0 idle, 0 running, 0 held


Submitting a job to Condor-G example 2

Run the (prestaged) Unix ls program on a different Globus resource:

executable = /bin/ls Transfer_Executable = false globusscheduler = vulture.cs.wisc.edu/jobmanager universe = globus output = ls-test.out log = ls-test.log queue

• The executable is pre-staged. Being on the remote machine, there is no need to transfer it before execution.

• The required globusscheduler and universe commands are present. • The command Transfer_Executable = FALSE identifies the executable as

being pre-staged. In this case, the executable command gives the path to the executable on the remote machine.


Submitting a job to Condor-G example 3

Submit a Perl script to be run as a Condor job. The Perl script both lists and sets environment variables for a job.

• Save the following Perl script with the name env-test.pl, to be used as a Condor job

executable: #!/usr/bin/env perl foreach $key (sort keys(%ENV)) { print "$key = $ENV{$key}\n" } exit 0;

• Run the Unix command chmod 755 env-test.pl to make the Perl script executable.

• Create the following submit description file executable = env-test.plglobusscheduler = biron.cs.wisc.edu/jobmanageruniverse = globus environment = foo=bar; zot=qux output = env-test.out log = env-test.log queue



• When the job has completed, the output file env-test.out should contain something like this:

GLOBUS_GRAM_JOB_CONTACT = https://biron.cs.wisc.edu:36213/30905/1020633947/GLOBUS_GRAM_MYJOB_CONTACT = URLx-nexus://biron.cs.wisc.edu:36214 GLOBUS_LOCATION = /usr/local/globus GLOBUS_REMOTE_IO_URL =

/home/epaulson/.globus/.gass_cache/globus_gass_cache_1020633948 HOME = /home/epaulson LANG = en_US LOGNAME = epaulson X509_USER_PROXY = /home/epaulson/.globus/.gass_cache/globus_gass_cache_1020633951 foo = bar zot = qux



Of particular interest is GLOBUS_REMOTE_IO_URL environment variable: Condor-G automatically starts up a GASS remote I/O server on the

submitting machine. Because of the potential for either side of the connection to fail,

the URL for the server cannot be passed directly to the job. Instead, it is put into a file, and the GLOBUS_REMOTE_IO_URL

environment variable points to this file. Remote jobs can read this file and use the URL it contains to access the

remote GASS server running inside Condor-G. If the location of the GASS server changes (for example, if Condor-G

restarts), Condor-G will contact the Globus gatekeeper and update this file on the machine where the job is running.

It is therefore important that all accesses to the remote GASS server check this file for the latest location.


Submitting a job to Condor-G last example

A Perl script that uses the GASS server in Condor-G to copy input files to the execute machine. (the remote job counts the number of lines in a file.)

#!/usr/bin/env perl use FileHandle; use Cwd; STDOUT->autoflush(); $gassUrl = `cat $ENV{GLOBUS_REMOTE_IO_URL}`; chomp $gassUrl; $ENV{LD_LIBRARY_PATH} = $ENV{GLOBUS_LOCATION}. "/lib"; $urlCopy = $ENV{GLOBUS_LOCATION}."/bin/globus-url-copy"; # globus-url-copy needs a full pathname $pwd = getcwd(); print "$urlCopy $gassUrl/etc/hosts file://$pwd/temporary.hosts\n\n"; `$urlCopy $gassUrl/etc/hosts file://$pwd/temporary.hosts`; open(file, "temporary.hosts"); while(<file>) { print $_; } exit 0;


Submitting a job to Condor-G last example Cont.

• The submit file: executable = gass-example.pl globusscheduler = biron.cs.wisc.edu/jobmanager universe = globus output = gass.out log = gass.log queue

• There are two optional submit description file commands of note: x509userproxy and globusrsl.

1) The x509userproxy command specifies the path to an X.509 proxy, as: x509userproxy = /path/to/proxy

– If this optional command is not present in the submit description file,then Condor-G checks the value of the environment variable X509_USER_PROXY for the location of the proxy.

– If this environment variable is not present, then Condor-G looks for the proxy in the file /tmp/x509up_u0000, where the trailing zeros in this file name are replaced with the Unix user id.


Submitting a job to Condor-G last example Cont.

2)The globusrsl command is used to add additional attribute settings to a job's RSL string, as:

globusrsl = (name=value)(name=value)

An example of this command in a submit description file

globusrsl = (project=Test_Project)

This example's attribute name for the additional RSL is project, and the value assigned is Test_Project.


Limitations of Condor-G

• No checkpoints. • No matchmaking. • File transfer is limited. There are no file transfer

mechanisms for files other than the executable, stdin, stdout, and stderr.

• No job exit codes. Job exit codes are not available.

• Limited platform availability. Condor-G is only available on Linux, Solaris, Digital UNIX, and IRIX. HP-UX support will hopefully be available later.


References Globus Project www.globus.org

• Overviews of Grid computing:Anatomy of the grid: http://www-fp.globus.org/research/papers.html#anatomyPhysiology of the grid:

http://www-fp.globus.org/research/papers.html#OGSA Older, extensive:

The Grid: Blueprint for a New Computing Infrastructure, I. Foster and C. Kesselman (Eds), Morgan Kaufmann, 1999.

• Globus FAQ http://www-fp.globus.org/about/faq/general.html• Globus installation http://www-fp.globus.org/gt2/admin/guide-verify.html• Condor-G manual:

http://www.cs.wisc.edu/condor/manual/v6.4/5_2Condor_G.html• A topical school on Grid computing will be held in Vico Equense, Italy

during the last two weeks of July, 2003.For details, send an email to [email protected].

Global Grid Forum www.gridforum.org

http://www.cs.wisc.edu/condor/manual/v6.4/5_2Condor_G.html

Documents

Introduction to Globus with Condor-G