RMS
Reference Manual
Quadrics Ltd.
© Quadrics Ltd. 1998-2007
Edition 17 - August 17th 2007
Contents
Chapter 1 - Introduction
Scope of Manual
Audience
Using this Manual
Related Information
Location of Online Documentation
Reader's Comments
Conventions
Chapter 2 - Overview of
RMS
Introduction
The System Architecture
The Role of the
RMS
RMS
Management Functions
Chapter 3 - Parallel Programs Under RMS
Introduction
Resource Requests
Loading and Running Programs
Chapter 4 -
RMS
Daemons
Introduction
The Database Server
The Machine Manager
The Partition Manager
The Switch Manager
The Transaction Log Manager
The Event Manager
The Process Manager
The RMS Daemon
Chapter 5 -
RMS
Commands
allocate
msqladmin
nodestatus
prun
pps
rcontrol
rinfo
rms
rmsarchive
rmsbackup
rmsbuild
rmsctl
rmsexec
rmshost
rmspost
rmsquery
rmstbladm
swctrl
Chapter 6 - Access Control, Usage Limits and Accounting
Introduction
Users and Projects
Access Controls
How Access Controls are Applied
Other UNIX Limits
Accounting
Chapter 7 -
RMS
Scheduling
Introduction
Scheduling Policies
Scheduling Constraints
What Happens When a Request is Received
Memory Limits
Swap Space
Draining Jobs
Timeslicing
Suspend and Resume
Idle Time
Complex Resource Requests
Chapter 8 - Event Handling
Introduction
Event Handling
List of Events Generated
Extending the
RMS
Event Handling Mechanism
Chapter 9 - Setting up RMS
Introduction
Installation Planning
Installing RMS
Upgrading to a new version of RMS
Setting up
RMS
Inter-Process Communication in
RMS
Local Customisation of
RMS
Setting up SSH
Setting up Kerberos Support
Time Synchronisation
The
RMS
cron
job
Setting up Timeslicing
CPUSET Support
Integration with the HP XC Cluster Software
Chapter 10 - RMS License Management
Introduction
Downloading the Software
Running
RMS
with a Demo License
Running
RMS
with a Permanent License
Getting Information on RMS License Management
Running
RMS
with a Node Locked License
Troubleshooting RMS License Management
Chapter 11 - Day-to-Day RMS Operation
Introduction
Periodic Shift Changes
Summarising Accounting Data
Archiving Data from the
RMS
database
Backing Up the
RMS
Database
Restoring the Database
Troubleshooting
Configuring Nodes Out
Cleanup
Draining Jobs From Nodes
Draining Jobs From Rails
Log Files
Chapter 12 - The RMS Database
Introduction
General Information about the Tables
Table Definition Files
Categories of Table
Listing of Tables
Appendix A - QsNet Terms
Introduction
Capabilities
Network Topology
Database Representation
Control Interface
Link States
Link Errors
Appendix B - RMS Status Values
Overview
Generic Status Values
Job Status Values
Link Status Values
Switch Network Status Values
Environment Status Values
Node Status Values
Partition Status Values
Resource Status Values
Transaction Status Values
Appendix C - RMS Kernel Module
Introduction
Capabilities
System Call Interface
rms_setcorepath
rms_getcorepath
rms_prgcreate
rms_prgdestroy
rms_prgids
rms_prginfo
rms_getprgid
rms_prgsuspend
rms_prgresume
rms_prgsignal
rms_prgaddcap
rms_setcap
rms_ncaps
rms_getcap
rms_prggetstats
rms_setpset
rms_getpset
rms_elaninitdone
The
RMS
/proc
interface
Appendix D - RMS Application Interface
Introduction
Manual pages
rms_allocateCpus
rms_allocateHosts
rms_allocateResource
rms_attach
rms_detach
rms_closeDatabase
rms_createResource
rms_deallocateResource
rms_defaultPartition
rms_numCpus
rms_numNodes
rms_machineName
rms_partitionIndex
rms_whichPartition
rms_errorString
rms_freeCpus
rms_freeCpusPerNode
rms_getNodeStats
rms_jobIds
rms_killJob
rms_nodeInfo
rms_partitionInfo
rms_resourceId
rms_resourceStats
rms_run
rms_suspendResource
rms_resumeResource
rms_killResource
rms_setResourcePriority
rms_version
rms_checkVersion
Appendix E - Accounting Summary Script
Introduction
Command Line Interface
Example Output
Listing of the Script
Glossary
Index
© Quadrics Ltd. 1998-2007