C H A P T E R  7

Displaying Program Information With mpps


What You Can Do

To Perform This Task

Use this Option

How to display job status

none

How to display information about individual jobs

-J

How to display name, PID, and host of current job

-b

How to display information about all jobs

-e

How to display a job's start time

-f

How to display job information by partition

-A -a

How to display job information by process

-P -p


 

 

 


procedure icon  How to Display Job Status

To display status information about your jobs running in the default partition, enter the mpps command without options:

% mpps

For example:

% mpps
JOBNAME   NPROC  UID   STATE  AOUT
cre.41    3      slu   RUN    AAA
cre.46    4      slu   EXNG   tmp
cre.49    1      slu   EXIT   tmp
cre.99    9      slu   EXNG   uname
cre.100   9      slu   EXNG   uname

The status fields are described in TABLE 7-1.

TABLE 7-1 Job Status Displayed by mpps

mpps Output

Description

CORE

The job or process exited due to a signal and core was dumped.

CORING

The job is exiting due to a signal. The first process to die dumped core.

EXIT

The job or process exited normally.

EXITING

The job is exiting. At least one process exited normally.

FAIL

The job or process failed while starting, or was aborted.

FAILING

Initialization of the job failed, or a job abort has been signaled.

RUN

The job or process is running.

SIGNALING

The job or process exited due to a signal.

SIGNALED

The job is exiting due to a signal. The first process to die was killed by a signal. At least one of its processes is still in the RUN state.

SPAWN

The job or process is being spawned.

STOPPED

The job or process is stopped.


 

 


procedure icon  How to Display Information About Individual
Jobs (-J)

To display information about a job, use the -J option and a job-attribute.

% mpps -J job-attribute[,job-attribute...]

Separate multiple job-attributes either with a comma or a space, but not both.

TABLE 7-2 Job attributes for -J option to mpps

Attribute

Description

part

The name of the partition running the job

jobname

The job's unique ID, expressed as <resource-manager>.jobname

mprun_pid

The process identifier (PID) of the current mprun job

mprun_host

The host of the current mprun job

nproc

The number of processes requested (the actual number of processes started may differ if the -W (How to Wrap Multiple Processes (-W)) or -S (How to Settle for Available Processes (-S)) flags were used with mprun)

uid

The user on whose behalf the job was run (normally the user who submitted the job)

gid

The group on whose behalf the job was run (normally the group of the user who submitted the job)

state

BUILD - The job is being submitted

WAIT - The job is waiting to run

SPAWN - The job is preparing to run

RUN - The job is running

RSTRT - The job has been killed because one of the nodes on which it was running went down; the job will be restarted

running

The number of processes actually running in this job. Not always equal to the number of processes started for the job because processes that have exited are not counted

wkdir

The directory in which the job's processes start

aout

The name of the program

paout

The full path of the program

ctime

The time when mprun was invoked

args

The command-line arguments of the program

stime

The time the job was started

prio

The job priority (higher numbers run first)



procedure icon  How to Display Job Name, PID, and Host of Current Job (-b)

Use the -b option to display job name, process identifier, and host of a current MPI job.

% mpps -b


procedure icon  How to Display Information About All Jobs (-e)

Use the -e option to display information about all jobs.

% mpps -e


procedure icon  How to Display a Job's Start Time (-f)

Use the -f option to display the start time for each job.

% mpps -f


procedure icon  How to Display Job Information by Partition
(-A -a)

To display information about jobs running in all partitions, use the -A option.

% mpps -A

To display information about jobs running in a specific partition, use the -a option, followed by the name of the partition.

% mpps -a partition-name

 


procedure icon  How to Display Job Information by Process
(-p -P)

Use the -p option to include information about the processes that make up the jobs:

% mpps -p

For example:

% mpps -p
JID       NPROC  UID   STATE  AOUT   RANK  PID    STATE  NODE
lsf.2320  4      shaw  RUN    sleep  0     10190  RUN    node6
                                     1      4744  RUN    node7
                                     2     16564  RUN    node4
                                     3      9412  RUN    node5

The output fields are described in TABLE 7-3, below.

To display information about a particular process attribute, use the -P option:

% mpps -P process-attribute[,process-attribute...]

Separate multiple process-attributes either with a comma or a space, but not both. Use the attributes described in TABLE 7-3, below.

TABLE 7-3 Process attributes for -P option to mpps

Attribute

Description

rank

The rank of the process within the job

pid

The process ID

state

The current execution state of the process

iod

The process ID of the I/O daemon for this process

load

The load on the node executing the process

node

The name of the node executing the process



Command Reference (mpps)

TABLE 7-4 Options for mpps

Option

Description

none

Display status information about your jobs running in the default partition

-J

Display information about a particular job

-b

Display name, process identifier, and host of current job

-e

Display information about all jobs

-f

Display the time a job started

-A

Display information about jobs running in all partitions

-a

Display information about jobs running in a particular partition

-P

Display process information about a job

-p

Display information about a particular process attribute