Standards » Management Area » UR WG

h1. List of fields considered for compute accounting

h2. Processors

The number of processors used or requested. A processor definition may be dependent on the  machine  

architecture. Typically processor  is equivalent to the  number  of  physical CPUs used. For example, if a 

job uses two cluster “nodes”, each node having 16 CPUs each, the total number of processors would be 

32.

Example

h3. Comments:

Ralph: What about cores? Do we need to differentiate between cores and processors? 

----

h2. TimeDuration

This property identifies any additional measure  of  time  duration associated with the

resource  consumption. For  example, it may report the  connection time  within a multitasking queue.

Example

h3. Comments:

Ralph: How is it expressed what kind of time duration this property refers to? An attribute? 

Jon: Wasn't it decided to use StartTime and EndTime instead of TimeDuration?

----

h2. TimeInstant

This property identifies any additionally identified discreet timestamp associated with the

resource consumption. For example, it may represent the time the job was queued, rather

than the start-time of the job.

Example

h3. Comments:

Ralph: How is it expressed what kind of time this property refers to?  An attribute? 

---

h2. WallDuration

This element specifies the wall clock time that elapsed. This optional attribute provides additional information about the wall clock time reported.

Example

h3. Comments

Jon: Do we need this?

---

----

h2. ALL OK:

----

h2. ComputeUsageBlock

This is the block property of the compute usage fields. Similar blocks for storage and other resources should also be added.

Example

<pre>

<ur:ComputeUsageBlock>

  <!—Compute Record properties go in here -->

</ur:ComputeUsageBlock>

</pre>

h3. Comments:

Ralph: ok

----

h2. CpuDuration or ComputeDuration

CPU time used, summed over all processes in the job.

Example

h3. Comments:

Ralph: ok

----

h2. StartTime

The time at which the usage consumption (i.e. “job”) started.  The value of this property 

may depend on the selected queue system. For example, some systems include time to 

stage files, others do not.

Example

h3. Comments:

Ralph: ok

----

h2. EndTime

The time at which the usage consumption (i.e. “job”) completed. The value of this 

property may depend on the selected queue system. For example some systems include 

time to stage files, others do not.

Example

h3. Comments:

Ralph: ok

----

h2. MachineName

A descriptive name of the machine on which the job ran. This may be a system hostname, 

or may be a site’s name for a cluster of machines. The identification of the machine by 

name may assume the context of the grid in which the machine participates; i.e. machine 

names may be unique within a specific grid, but do not need to be unique across the set of 

all grids.

Example

h3. Comments:

Ralph: ok

----

h2. NodeCount

Number  of  nodes used. A node  definition may be  dependent on the  architecture, but

typically a  node is a physical machine. For  example  a  cluster  of  16 physical machines

with each machine  having one  processor  each is a  16 “node”  machine, each with one

“processor”. A 16 processor SMP machine however, is 1 physical node (machine) with

16 processors.

Example

h3. Comments:

Ralph: ok

----