[oslc-Metrics] new Metric and UnitOfMeasure subclasses

Tue Oct 2 11:45:50 EDT 2012

Julie,

I'd have to look at your design to see if you need new basic predicates. 
For example, why do you feel it's important to distinguish integers from 
doubles? A double can easily represent integers (and booleans as 0 and 1). 
The reason for having a single numeric datatype is simplicity and 
uniformity. 

Also, what you think is an integer might become a double if you change 
units. For example, lines of code (LOC) is an integer. Suppose a program 
has 2500 LOC. But suppose you chose KLOC as the unit of measurement. Then 
the numeric value is 2.5, a double.

Regards, 
___________________________________________________________________________ 

Arthur Ryman 

DE, Chief Architect, Reporting &
Portfolio Strategy and Management
IBM Software, Rational 

Toronto Lab | +1-905-413-3077 (office) | +1-416-939-5063 (mobile) 

From:
Julianne Bielski/Raleigh/IBM at IBMUS
To:
Arthur Ryman <ryman at ca.ibm.com>
Cc:
Janet Andersen/Raleigh/IBM at IBMUS, John Arwe/Poughkeepsie/IBM at IBMUS, 
oslc-metrics at open-services.net, oslc-metrics-bounces at open-services.net
Date:
09/26/2012 02:12 PM
Subject:
Re: [oslc-Metrics] new Metric and UnitOfMeasure subclasses

Arthur, 

Per: Also, for the new units, you can simply use the URIs that are defined 
in 
other vocabularies if they have the desired meaning. For example, the EMS 
spec recommends the use of terms defined in QUDT. 

If <ems:numericValue> is supposed to be of type double, but other metric 
values are integers (like counts), would I just create my own predicates 
for ems:Measure like pm:count, pm:boolean?

-- Regards,

Julianne Bielski, STSM
ITM Core Chief Architect
Tivoli, IBM Software Group
tel: (919) 224-1170      (T/L) 687-1170
e-mail: bielsk at us.ibm.com

"All growth is a leap in the dark, a spontaneous unpremeditated act 
without benefit of experience." ? Henry Miller 

From:        Arthur Ryman <ryman at ca.ibm.com> 
To:        Julianne Bielski/Raleigh/IBM at IBMUS, 
Cc:        Janet Andersen/Raleigh/IBM at IBMUS, John 
Arwe/Poughkeepsie/IBM at IBMUS, oslc-metrics at open-services.net, 
oslc-metrics-bounces at open-services.net 
Date:        09/19/2012 03:58 PM 
Subject:        Re: [oslc-Metrics] new Metric and UnitOfMeasure subclasses 

Julie,

I've looked at your proposal. It may be better to put these in a 
vocabulary defined specifically for Performance Monitoring and owned by 
your workgroup. There is no requirement that all metrics belong to an EMS 
vocabulary. Your workgroup could create a subclass of ems:Metric.

e.g. use the vocabulary URI <http://open-services.net/ns/perfmon/metric#>, 

and the prefix pmm:

pmm:Metric rdfs:subClassOf ems:Metric

And make your new metrics members of this class, e.g.

pmm:FreeMegabytes a pmm:Metric .

Also, for the new units, you can simply use the URIs that are defined in 
other vocabularies if they have the desired meaning. For example, the EMS 
spec recommends the use of terms defined in QUDT.

Regards, 
___________________________________________________________________________ 

Arthur Ryman 

DE, Chief Architect, Reporting &
Portfolio Strategy and Management
IBM Software, Rational 

Toronto Lab | +1-905-413-3077 (office) | +1-416-939-5063 (mobile) 

From:
Julianne Bielski <bielsk at us.ibm.com>
To:
oslc-metrics at open-services.net
Cc:
Janet Andersen <janetand at us.ibm.com>, John Arwe <johnarwe at us.ibm.com>
Date:
09/19/2012 03:30 PM
Subject:
[oslc-Metrics] new Metric and UnitOfMeasure subclasses
Sent by:
oslc-metrics-bounces at open-services.net

OSLC EMS workgroup members, 

The Performance Monitoring domain specification members would like to 
propose the set of metrics and units below to be added to the ems spec. 
Could you please review and respond with any questions and hopefully 
agreement by 9/24? 
Performance Monitoring Metrics 
Performance Monitoring metrics are metrics that can be measured on a 
resource and give an indication of the performance, availability, 
utilization, or capacity. The Performance Monitoring metric classes are 
subclasses of ems:Metric. 
Table of Performance Metric Standard URIs 
URI
Description
http://open-services.net/ns/ems/metric#FreeMegabytes 
Number of megabytes free on a logical disk. 
http://open-services.net/ns/ems/metric#VirtualMemoryUtilization 
Percentage of virtual memory in use by a computer system or process as a 
percentage of all available virtual memory. 
http://open-services.net/ns/ems/metric#RealMemoryUtilization 
Percentage of real memory in use by a computer system or process as a 
percentage of all available real memory. 
http://open-services.net/ns/ems/metric#CPUUtilization 
Percentage of CPU time used by a computer system or process as a 
percentage of total available CPU time. 
http://open-services.net/ns/ems/metric#ResponseTime 
Time it takes for a response to be returned to a request 
http://open-services.net/ns/ems/metric#NumFailedSqlStmts 
The number of failed SQL statements returned from a database to a client 
http://open-services.net/ns/ems/metric#HeapUsage 
The amount of memory heap used by an application 
http://open-services.net/ns/ems/metric#TableReorgNeeded 
Indicates whether a database?s tables need to be reorganized 
http://open-services.net/ns/ems/metric#PctBufferPoolUsed 
Percentage of buffer pool used by a database server 
http://open-services.net/ns/ems/metric#GarbageCollectionCount 
The number of times GC has been run during the lifetime of an application 
http://open-services.net/ns/ems/metric#PercentageFreeMemoryAfterGC 
Percentage of free memory on application JVM?s heap following a garbage 
collection. Indicator of memory leak or misconfiguration. 
http://open-services.net/ns/ems/metric#PercentageTimeThreadPoolMaxed 
Represents the proportion of time an application?s thread pool has been 
totally consumed since application start time. Indicator of thread pool 
contention and slowing response times 
http://open-services.net/ns/ems/metric#PercentageTimeJCAThreadPoolMaxed 
Represents the proportion of time a JCA application?s thread pool has been 

totally consumed since application start time. Indicator of thread pool 
contention and slowing response times 
http://open-services.net/ns/ems/metric#PercentageTimeDBThreadPoolMaxed 
Represents the proportion of time a database application?s thread pool has 

been totally consumed since application start time. Indicator of thread 
pool contention and slowing response times 
http://open-services.net/ns/ems/metric#VMCPUPercentReady 
The amount of time a Virtual Machine?s CPU has spent in the ready state. A 

relatively high number indicates under utilization 
http://open-services.net/ns/ems/metric#TotalCPUAvailable 
The amount of MHz available to a physical server 
http://open-services.net/ns/ems/metric#OverCommmittedDiskUtilization 
The percentage of the total capacity of a storage volume that has been 
overcommitted.

Performance Monitoring Metrics Units of Measure 
The Performance Monitoring metric units of measure are a subclass of 
ems:UnitOfMeasure. 
Table of ems:Unit of Measure Standard URIs for Performance Metrics 
URI
Description
http://open-services.net/ns/ems/unit#Percentage 
dbpedia:Percentage 
http://open-services.net/ns/ems/unit#Megabytes 
qudt:Capacity 
http://open-services.net/ns/ems/unit#MegaHertz 
qudt:MegaHertz

-- Regards,

Julianne Bielski, STSM
ITM Core Chief Architect
Tivoli, IBM Software Group
tel: (919) 224-1170      (T/L) 687-1170
e-mail: bielsk at us.ibm.com

"All growth is a leap in the dark, a spontaneous unpremeditated act 
without benefit of experience." ? Henry Miller
_______________________________________________
Oslc-Metrics mailing list
Oslc-Metrics at open-services.net
http://open-services.net/mailman/listinfo/oslc-metrics_open-services.net