IBM Support

How to understand peformance metrics generated by the Netcool Omnibus Java Remedy Gateway?

Technical Blog Post


Abstract

How to understand peformance metrics generated by the Netcool Omnibus Java Remedy Gateway?

Body

 

 

How to understand performance metrics generated by the Netcool Java Remedy Gateways?

 


Synopsis : Users of the Java Remedy Gateway may want to check how well the gateway is performing by viewing the STATISTICAL metrics in the gateway log files
                        
here is an example of what the stats outputs look like in the java remedy gateway log file

17/02/24 12:03:08: Information: [Main Gateway] STATS: [IDUC processing] count=2781 time=956 rate=2908.9958 rolling=176.47058
17/02/24 12:03:08: Information: [Main Gateway] STATS: [IDUC processing] count=2781 uptime=65 rate=42.46061 busy=1.4596311%
17/02/24 12:03:08: Information: [SAF Processor] STATS: [SAF processing] count=188 time=511 rate=367.90607 rolling=367.90607
17/02/24 12:03:08: Information: [SAF Processor] STATS: [SAF processing] count=188 uptime=61 rate=3.043795 busy=0.8273294%


                                                                             
Note that the performance metrics reported by the gateway are in fact target system independent,
and are generated by the generic framework  rather than the Remedy specific portion, and so the same metric                 
information applies to the ServiceNow gateway as well.)                         
                                                                                
Taking the following example

 

17/02/24 23:56:35: Information: [SAF Processor] STATS: [SAF processing]  count=170 time=425 rate=400.0 rolling=400.0  
17/02/24 23:56:35: Information: [SAF Processor] STATS: [SAF processing] count=170 uptime=1 rate=114.01744 busy=28.50436%


lets look at each line in turn:                                                       
                                                                                
17/02/24 23:56:35: Information: [SAF Processor] STATS: [SAF processing]  count=170 time=425 rate=400.0 rolling=400.0                                     
                                                                                
- count is the total number of datums processed in SAF. A datum can be an alert, a journal or a detail.                                                
- time is the time it has taken to process those datums (in ms.)                
- rate is the rate of processing of the last SAF batch of data  (datums/second)                                                                 
- rolling is the rolling average rate of processing over the last batches.

 

Historical rolling rate is blended with the last batch rate to produce a new rolling rate.
So the rolling rate is the metric we're aiming at monitoring for general over time performance.                                
                                                                                
17/02/24 23:56:35: Information: [SAF Processor] STATS: [SAF processing] count=170 uptime=1 rate=114.01744 busy=28.50436%                                
                                                                                
- count is as before, the total number of datums.                               
- uptime is the number of milli-seconds the gateway has been running            
- rate is the aggregate lifetime rate of processing of alerts. This will usually be much lower than the rolling rate recorded previously, as it          
  also includes idle time when no alerts are being processed.                     
- busy is the percentage of time the gateway spends actually processing SAF.

 

So the lower here the better.

 

If it is

 

> 70%  then the gateway is pretty busy, and if it is

> 90%  we are starting to get close to  the limits of what we can handle into the target,

 

and should perhaps look at increasing the connection count in the gateway props file

 

The default location and name specified in the G_BMC_REMEDY.props properties file is '$OMNIHOME/etc/G_BMC_REMEDY.props'.

 

Gate.Remedy.Connections integer

-remedyconnections integer

Use this property to set the number of simultaneous connections the gateway makes with BMC Remedy ARS.

The default value is 2.

 

                            
                                                                                
The SAF STATS are most pertinent to the performance of the gateway, and indicates how well the gateway is writing to the target system.

If the busy percentage is high, or the gateway is backing up in SAF processing(which will be indicated by an increase in the number of files in the gateway data directory and a corresponding GatewayWatch alert being  created in OMNIbus) then it is an indication it might be beneficial to  increase the connection count to the target to increase concurrency.

Up to the limits of the target Remedy server, performance should increase fairly linearly with the number of connections, and Remedy systems with         
complex workflows will benefit the most from a large connection count.          

Here are some further notes on SAF file processing should your require further info

https://developer.ibm.com/answers/questions/289903/how-does-the-ibm-tivoli-netcoolomnibus-java-gatewa/#
                                                                                
The IDUC STATS lines have identical metrics, but IDUC processing  generally isn't a bottleneck and so is of less importance. But if the busy percentage of IDUC processing is high, we have less scope to increase performance to compensate without increasing CPU and/or IO  performance, as IDUC processing is inherently single threaded.    

 

 

Subscribe and follow us for all the latest information directly on your social feeds:

 

 

 

image

 

 

image

 

 

image

  

 

Check out all our other posts and updates:

Academy Blogs:https://goo.gl/eZjStB
Academy Videos:https://goo.gl/BMc9hg
Academy Google+:https://goo.gl/ukxTcP
Academy Twitter :https://goo.gl/kQFMpN

 

 

image


 

IBM Netcool Support Home Page and Links to Knowledge Center 

https://www.ibm.com/support/knowledgecenter/SSSHTQ/landingpage/NetcoolOMNIbus.html

http://ibm.co/29S5utj

[{"Business Unit":{"code":"BU053","label":"Cloud & Data Platform"},"Product":{"code":"","label":""},"Component":"","Platform":[{"code":"","label":""}],"Version":"","Edition":"","Line of Business":{"code":"","label":""}}]

UID

ibm11081437