onTune the differences

onTune SPI
Distinctions from The Others

Contents
onTune distinctions
Case Study3
Comparison data with other solutions2
1

 Collecting & providing system performance data with realtime
 Providing data chart with realtime
 Collecting & providing event data with realtime
 Notifying alert to users with realtime
1 Realtime performance monitoring
 Collecting the performance data of all the system resources
 All the system data collected without customizing
 Extra data collected through customizing
2 Collecting various performance data
 Lookup for past data in single screens
 Available to apply same data to every chart
 Keeping past performance data by second
3 Lookup for past data
 By providing quick & accurate data, the causes of issues are
revealed
 Storing all the data, users can get data needed in anytime &
anywhere
4 Handling issues
 Every chart and data on the screen can be documented
 Reporting tool provided
 Easy to use with promptness
5 Reporting tool
 VIO, Micro Partition monitoring
6 IBM virtualization supporting
No solution available for realtime monitoring
The others need customizing according to policy
It is impossible for 5min, 10min data to provide
distinctive data
No solution available to monitor IBM virtualization
Not available for the others to save past data by
second unit
It takes huge time and effort for the others to make
performance data

1. Realtime Monitoring
The strongest feature OnTune has is realtime monitoring by second, which the other solutions are not able to support. With the feature,
OnTune provides more accurate and prompt data than any other solutions provide.
In addition to performance analysis, when issues happen, OnTune provides proper data so that users can analyze accurate causes of issues.
Even event occurrence is notified to users with realtime so that OnTune can inform issues to the users sooner than the other solutions
Shows system information by
seconds

1 minute
10 minutes
What make a difference!
2 seconds
* The difference between second-unit monitoring and minute-unit monitoring

2. Collecting various performance data
OnTune collects currently all the system performance data without any customizing. So, it supports analysis unexpected bottleneck or
performance issues.
Because all the data of resources can be compared and analyzed even at the time issues happen, OnTune can provide the most accurate data
to know the causes of issues.
Without extra customizing,
performance data of system
resources being composed &
collected automatically.

3. Lookup for past data
It is easy for OnTune to analyze past data because it stores the data of last time by senconds.
Above all, it contribute a convenient feature which past data and current data are compared on a single screen so that users can analyze
easily.
Choose a certain time of
past
The realtime graph is
being continually marked
The past graph emerges

4. Handling issues
When a issue happens, operators need data as much as possible to figure out the causes. OnTune is the best alternative.
It stores every log occurring in system resources, and the data is stored by seconds so that even very short time issue can be delivered to
operators.
Ontune sends alerts about event with realtime, so it notifies operators events occurring in system more promptly and accurately than any
other solutions
OnTune also stores long-term data , so users can confirm the subtle changes for a long period. This is a crucial role for analyzing issues and
finding the causes of issues.
With long-term data, the
suspicious time for issues
can be checked
Check the detailed time
Various items can be
compared in a single screen

5. Reporting Tool
onTune reporting tool helps operators to save time and efforts for making performance data , so it is available to invest the time and effort to
the other valuable things
All the data and graphs shown on the screen can be stored as xls or ppt file through OnTune.
The strong and intuitive reporting tool makes operators save time a lot, and help them to watch performance data and to make reports easily
with simple control in case of emergency.
Actually, system operators spend lots of time for making reports. However, OnTune reduces the time for reports dramatically so that operators
can focus on the valuable business.

Function OnTune Patrol HP OpenView IBM Tivoli
Real time monitoring by the second O

Log collecting interval getting
longer if objects are
numerous

Log collecting interval
getting longer if objects
are numerous

Log collecting cycles
getting longer if
objects are numerous
OS parameter monitoring2) O X X X
Configuration of monitoring interval
individually by system groups3)
O O O O
Creating performance statistic data
without limit time interval(second ~
year)
O X X X
Performance reports available4) O

Customizing needed
(High cost)

Customizing needed
(High cost)

Customizing needed
(High cost)
Less resources consumption for
Agents5)
(CPU: less1 than 1%, Memory: Less
than 20Mbyte)
O
X
Depends on the objects’
number
X
number
X
number
Infra expense for installation Low
High
Customizing needed in most
cases
High
Customizing needed in
most cases
High
Customizing needed in
most cases
onTune vs. BMC Patrol / HP OpenView / IBM Tivoli (I)
1) Available to monitor objects in the second- Generally other solutions take 30sec~ few minuet for monitoring, and the interval get longer as objects increase in number
2) Easy to confirm the factors affecting performance of servers by managing OS configuration value (OS parameter)
3) Available to make groups with business service or productions/developers/tests in terms of servers, and set monitoring interval for the each group individually
4) Available to make performance or issues reports for days, weeks, and months through excel and PPT with simple clicks.
5) Despite of performing real time monitoring in the second, Ontune shows that agents use only small part of CPUs.

Functions OnTune Patrol HP OpenView IBM Tivoli
Monitoring CPU/Memory/IO of PID6)
O X X X
Monitoring CPU/Memory/IO of a process by Command7)
O X X X
Monitoring CPU/Memory/IO of a process by Users8)
O X X X
Monitoring Memory Leak of a process9)
O   
PSMD Monitoring10)
O X X X
Monitoring for usage of individual Adapters (Network, FC
Adapter)
O X X X
Monitoring usage of individual physical disks O X X X
Monitoring for Volume Groups and Assigned disks O X X X
Monitoring Veritas Volume Manager O X X X
Easy monitoring and analysis, so training is not needed
O
Intuitive interface, ease
to use
X
Special training needed
for complex UI
X
Special training needed for
complex UI
X
Special training needed for
complex UI
IBM virtualization monitoring11)
O X X X
onTune vs. BMC Patrol / HP OpenView / IBM Tivoli (II)
6) Once agents are installed , Ontune performs monitoring for performance resources usage (CPU / Memory / IO) of all processes without extra setting.
7) Once agents are installed , Ontune performs monitoring for performance resources usage (CPU / Memory / IO) according to users’ command groups ( i.e. All the java processes
of webwas users) without extra setting.
8) Once agents are installed , Ontune performs monitoring for performance resources usage (CPU / Memory / IO) according to users (i.e. accounts of -was, root, or oracle) without
extra setting.
9) As a function explained in 6), when memory leakage is suspicious in a certain server, it is possible to diagnose which process makes trouble and when it is started through history
analysis about memory usage of each process.
10) Available to monitor Promotion/Demotion of IBM AIX’s Multi Page Memory Pool (4K/64K)
11) Available to monitor CPU/Memory Pool , VIO’s performance for IBM virtualized servers

OnTune Case Study
Migrating servers from HP to IBM, CPU usage increased abnormally.
Case study as a breakthrough for performance issue1
OnTune discovered the CPU reached the peak every 20 seconds showing letter ‘M’, and we notified that to the
developers so that they can realize hinit and fix the logic.
Large scale business like migration desperately needs monitoring on systems and performance.
Without OnTune monitoring by second, the pattern explained above cannot be recognized. If possible, it takes many
hours or days to collect log files, convert and analyze data.
This case is a proper example proving that OnTune provides realtime data to diagnose issues immediately.

Short term performance degradation emerged on a DB server
Monitoring critical system and confirming issues2
OnTune figured out the system call and interrupt decreased rapidly in 2~3 seconds and the symptom repeated every
minuet. Based on the data OnTune provided, the company fixed the bug between Hypervisor and OS on
communication.
The other solutions are not able to collect the data such as system call or interrupt, so those solutions have trouble
with confirm the cause of performance degradation.
Moreover, this kind of log files occurring instantly cannot be caught by other solutions which are impossible to
monitor with realtime.
OnTune Case Study

Finding out the memory leakage of ERP system
Memory leakage for long term3
Memory of a system increased 30% consistently for 3 months. The operators figured out the certain process causing
memory increase through OnTune’s long-term data.
OnTune helps users to analyze long-term progress, keeping data of each process.
The case explained above is a good example as prevention keeping every detailed data for long-term.
OnTune Case Study

Primary system crash occurred and then secondary system crash in Oracle DB composed with H/A system.
Confirming the issues of High Availability System4
By comparing system data of both side, OnTune confirmed the memory usage of demon monitoring dual system
had been increasing right before the issue emerged, and the pattern of the issue occurring in specific time ( when
the process memory reaches 1.2GB). Based on the data, users could check the bug and fixed it.
The data can be checked easily after issues happen because OnTune collects all the data about all the resources of
system.
The case indicates that OnTune checked memory data of PID until right before the issue happened so that it can
analyze the cause of issues.
OnTune Case Study

Oracle server showed sudden time-out and process delay
Case of Time-out occurred in DB server5
Users compared and analyzed the usage and throughput of individual disks via OnTune, and then discovered the disk
busy increased despite of little I/O amount. In addition, they confirmed the composition of external disk and realized
the issue was caused by I/O occurring in LUN of physical disk
This case proves that OnTune analyzes each disk’s IO/Busy in a single chart containing many servers.
OnTune Case Study

CPU performance issue in composition of Micro-Partition
A CPU performance issue in IBM’s Micro-Partition6
A certain partition of MP showed CPU performance degradation. With onTune monitoring CPU pool, users confirmed
that the certain partition interrupted other partitions not to be assigned resources. Based on the data, resource
distribution was adjusted by the users.
OnTune is the only solution available to monitor IBM’s Micro-Partition , so the other solutions are not able to solve
the issue explained above, even monitoring VIOS is impossible for them.
Through the function, OnTune can check efficient CPU resource and usage.
OnTune Case Study

After replacing old system, confirming tuning point of system configuration
System configuration issue after replacing legacy system.7
After migration, OS parameter was initialized, so users needed the optimal configuration value for the OS parameter.
Because OnTune keeps system parameter regularly, users can decide the proper OS parameter for the new system
by comparing previous one and other system.
OnTune provides lots of convenient functions to operators. Amid it keeps system parameter, users can check if the
set point changes or not.
Moreover, in case of adding similar systems, users can define standard configuration by confirming other system
configuration.
OnTune Case Study

Case of sudden DB process down in Sybase DB
DB Process went down in huge data DB8
With OnTune confirming the data of the issue time, users realized a log analysis process suddenly occupied Memory
to use, and system closed unstably with the amount of Memory lost by a DB process.
After notifying the Owner of analysis process and time to the operator, the process was moved to other system.
As a case of OnTune finding the issue by comparing each resource’s usage, the exact time and resources’ move can
be compared at the same time, and also even the Owner of each process and execution option can be checked.
Therefore, system operators are able to solve issues in short time.
In addition, migration considering resources of systems is possible because users can know the amount of resources
which processes need.
OnTune Case Study

System monitoring available for the large-sized HP system (94core, MEM 500GB, Disk 20000EA) by OnTune
Monitoring a large-sized system for new launching9
System resources getting more multiple and larger, the existing dash-boards or SMS can’t provide monitoring.
However, it is possible for OnTune to monitor the systems intensively by installing Agents only without any
configuration or customizing.
In the case of opening the large-sized system, the existing monitoring system can’t monitor all the resources with
defined items by system. In the other hands, OnTune doesn’t need to do extra customizing , and it can monitor every
single resource of system with simple agent installation. Therefore, the launching new system has been done
successfully with OnTune monitoring critical data and performance indicators.
OnTune Case Study

onTune the differences

More Related Content

What's hot (19)

Viewers also liked (9)

Similar to onTune the differences (20)

Recently uploaded (20)

onTune the differences

Editor's Notes