ibm operations analytics for z systems v2.2 - client long pres
TRANSCRIPT
© 2015 IBM Corporation
z Systems Software
IBM Operations Analytics for z Systems
Transforming Data into InsightsThe Next Generation of IT Service
Management
© 2015 IBM Corporation
z Systems Software
Note to the presenter …
This deck contains SEVERAL slides. It is intended to be modified by YOU to include the slides that YOU need for YOUR specific customer presentation.
If you are just looking for the high level overview of IOAz (‘short’ deck) OR the 1-pager slide, please retrieve those from the sales kit:
http://w3.ibm.com/software/spcn/content/F080197Y39289C99.html
2
© 2015 IBM Corporation
z Systems SoftwareAgenda
Why IT Analytics? Overview of IBM Operations Analytics for z Systems
– Functional capability– What’s New in 2015– Architecture– Out-of-the-box Value– Customize to meet your needs– Integration with Service Management tooling
Additional Detail – Bring Your own Data – Example using HMC log– IOAz V2.2 Details
– CICS insights– Network insights– Security insights– Log Forwarder improvements
3
© 2015 IBM Corporation
z Systems SoftwareSolution Branding
4
This solution was previously branded as IBM SmartCloud Analytics - Log Analysis.
The support to search and analyze z/OS logs was initially provided in March, 2014 under the following product names: IBM SmartCloud Analytics - Log Analysis z/OS - Insight Packs – SYSLOG V1.1IBM SmartCloud Analytics - Log Analysis z/OS - Insight Packs - IBM WebSphere® Application Server V1.1
Subsequent releases were named with the SmartCloud brand until April 2015, when Version 2 of the product was rebranded to
IBM Operations Analytics for z Systems
Initial release under the new name:IBM Operations Analytics for z Systems v2.1 (GA on April 24, 2015)
Current release:IBM Operations Analytics for z Systems v2.2 (GA on October 16, 2015)
Note that the distributed version of the product is now named IBM Operations Analytics – Log Analysis
© 2015 IBM Corporation
z Systems Software
5
z Systems scaling model and security to manage and optimize both
Systems of Record
Social, Mobile, Analytics Smarter Infrastructure
Business Transactions Quality of Service Command & Control Facts and data “source of truth” z/OS Systems
Mobile and Social Dynamic Interactions and
Collaboration Insight, trends, analytics
Systems of Engagement
Rapid growth of data from latest technologies can be supported seamlessly on z Systems
© 2015 IBM Corporation
z Systems Software
Analytics for System z addresses rapid growth of data and next generation technology
6
Much greater amount of critical IT operational data (SMF, log, journal) than distributed-only environments.
– Focus on problem determination and time to resolution while placing premium on availability of services and applications.
– 100x to 1000x explosion in data flooding existing tools. – New runtimes, programming languages needing complex
instrumentation.
By 2016, 40% of Global 2000 enterprises will have IT operations analytics architecture in place, up from < 1% in 2014, looking to integrate across their enterprise to reduce outages (Gartner).
90% of the Fortune 1000 companies are running z and have ‘Systems of Record’ dependencies for transactional processing and data serving applications.
© 2015 IBM Corporation
z Systems Software
Is managing IT today like sipping from a fire hose?
New Technologies like cloud, mobile and big data already challenging current Enterprise tools
7
Too long to isolate, diagnose problems in applications and infrastructure.
– Complex application workloads span multiple platforms – Increasing amounts of IT data: Performance metrics,
events, infrastructure logs, application logs, configuration files, traces
Existing IT tools need additional data analysis capabilities to manage of Systems of Engagement
– 100x to 1000x explosion in data flooding existing tools. – New runtimes, programming languages needing
complex instrumentation.
Reactive analytics misses critical information leading to outages
– Need to move to a more proactive model– Analysing ALL information better for predicting
problems.
© 2015 IBM Corporation
z Systems Software
Predict: Pro-Active Outage Avoidance Predict problems before they occur Search & Analyze: Quickly search and analyze large volumes of data from a single search barPerform log and performance analysis while searchingCorrelate messages from multiple logs for end-to-end problem diagnosisOptimize: Improve performance across IT Infrastructure
IBM is focused on managing end-to-end analytics for improved performance and workload management
8
IBM Analytics solutions for z Systems
PredictIBM zAware
Proactive Outage Avoidance
Search & Analyze
IBM Operations Analytics for z Systems
Faster Problem Resolution
OptimizeIBM Capacity
Management Analytics (CMA)
Optimized Performance
© 2015 IBM Corporation
z Systems SoftwareIBM Operations Analytics for z Systems
9
Analyze various types of data (logs, metrics, events, trouble tickets) from multiple sources (mainframe and distributed)
Locate problems from system, configuration, software logs and performance metrics using rapid index search and pattern analysis
Isolate issues across various domains including OS, Middleware, applications, etc. Leverage Expert Advice via links to support documentation and operations notes
to resolve problems quickly Visualize search results with analytic tools to rapidly determine root cause Out-of-the-box analysis and insights for z/OS, WebSphere, DB2, CICS, IMS, MQ,
Network, Security as well as distributed systems Enable early error detection and broaden
scope of automation with event notifications Fully customizable to meet your needs
SEARCH
ANALYZE Launch to Support Doc
RESOLVE
INTEGRATE
Accelerate problem isolation and identification … Reduce mean time to repair
• Network insights• Security insights• Event notification• Hadoop support• Analysis of performance metrics (SMF real time Data Provider)• Integration with existing Service Management tooling (Automation,
Monitoring, Event and Incident Management) • Role-based access control• Multi-time zone support
in 2015
© 2015 IBM Corporation
z Systems Software
Applications Search
Mainframe z/OS
z/Linux
z/O
S L
ogFo
rwar
der
Log
File
Age
ntor
Logs
tash
SM
FR
eal-t
ime
Dat
a P
rovi
der
WAS SYSOUT
z/OS Syslog
CICS MSGUSR
Annotators
CICS EYULOG
…
Operations Analytics Server
WAS SYSPRINT
WAS SYSOUT
DB2
DB2 App
Syslog
Web Access Log
The IBM Operations Analytics server is installed on z System (or x System) running Linux (64 bit)
z/OS Insight Packs are installed on the IBM Operations Analytics server z/OS Log Forwarder / SMF Data Provider installed on each z/OS LPAR
where you want to provide Search and Analysis
WAS SYSPRINT
Indexers
Current/ Archive
Tier
HadoopTier
USS Log Files
Other Logs
10
Alert Actions
Index
Log
Script
SNMP
EIF
AlertsInsight Pack
(z/OS)Distributed
Systems Insight Packs
SMF Data
Gen
eric
Rec
eive
r
Other Logs
Net
Vie
w
Mes
sage
G
athe
rer
IBM Operations AnalyticsArchitecture and Flows
Joblogs
NetView Netlog
……
© 2015 IBM Corporation
z Systems Software
Search specific logs or ALL logs
Timeframe
Enter search string
Save My Search
Simple search interface EASY to customize
11
Quick Searches, Analysis,Annotations, Patterns, Expert Advice, Dashboards will populate the Navigation tree
Log data is analysed and insights are surfaced as you searchFind problems you didn’t know existed
© 2015 IBM Corporation
z Systems SoftwareEasy to use – Quick Search
Domain-specific ‘Quick Searches’ available out-of-the-box
or create and save your own Provided with every z/OS Insight Pack
Provided by subject matter experts, support teams and customers
Immediate value out of the box
Easy to modify or create and save your own
12
© 2015 IBM Corporation
z Systems Software
Dashboards, Information Links and Expert Advice
Visualize the data with Dashboards Quick links to additional information and
support documents. Provided with every Insight Pack Expert Advice to access white papers,
tech notes, APARs, etc. for faster problem resolution
Dashboard views created by subject matter experts, support teams and customers
Immediate value out of the box Easy to modify or create and save your
own
13
© 2015 IBM Corporation
z Systems Software
Quickly and easily access IBM Support Portal based Expert Advice from Log Analysis
Launch to Tech Note
All IBM support site documents that reference messages from search results
Search for expert advice with the click of a button
14
© 2015 IBM Corporation
z Systems SoftwareAnalyze logs as you Search
Insights are surfaced automatically as you search. Patterns are
surfaced based on the log type. Provided with every Insight Pack Logs are analysed automatically Log data is categorized by hostname,
data source, message type, message source, etc.
Patterns/Insights are surfaced to help you focus on the source of the problem.For example, log analysis automatically surfaces java exceptions in application logs.
Perform searches and analyse multiple logs, organized per the needs of your enterprise.
Create your own Insight Pack for any text logs with time stamps
15
© 2015 IBM Corporation
z Systems SoftwareSample dashboardView your log and metric data however you like
16
© 2015 IBM Corporation
z Systems Software
Out-of-the-box dashboards (Example – Display message counts and java exceptions)OR
Build Your Own Dashboard with the click of the mouse
Sample dashboardView your log and metric data however you like
© 2015 IBM Corporation
z Systems SoftwareVisualizing the DataSearch and Analyze SMF Data (New in 2015)
Analyze your SMF data AND your log data for a complete view of the enterprise.
18
CPU utilization, Working Set Size, Paging & IO Rates
© 2015 IBM Corporation
z Systems Software
Perform simple free-form searches using the standard set of search keywords and operators Build complex queries with range searches and DateMath functions To learn more, consult Online Help available from the Learn More → Search Bar → Search query
syntax menu:
BYOD – Bring your own Data – The z/OS Log Forwarder can be configured to forward your text logs to enable Search, Analysis, Dashboards and Expert advice.
BYOIP – Build your own Insight Pack BYOV – Build your own Views (Graphs, Charts and Dashboards)
Out-of-the-Box capabilities provide immediate value.Additionally, IOA can easily be tailored to your specific needs.
Create your own – Queries, Dashboards, Feeds
19
© 2015 IBM Corporation
z Systems Software
Customer ExperiencesLarge Insurance Company (29 hours down to 29 minutes)Experienced an application outage that resulted in the team working around the clock for 29 hours. Multiple customers and IBM support staff poured through logs and traces to determine the root cause of the issue. After the issue was resolved, the logs were captured and sent to IBM lab for analysis using IBM Operations Analytics for z Systems. Within minutes, the IBM team was able to focus in on the root cause of the problem and to find the relevant PTF to resolve the issue through the integrated expert advice.
State Agency (up and running in 2.5 hours)Were able to download, install, configure and use IBM Operations Analytics for z Systems to search their logs in 2.5 hours.
Numerous Customers (improve visibility and find problems you weren’t aware of)Errors lurking in logs that are never examined because they don’t necessarily cause SLA or performance problems. For example, IBM Operations Analytics for z Systems found Over 4,000 invalid login attempts in a three day period that had otherwise gone unnoticed. MQ channel errors causing MQ errors in logs from distributed systems – not being monitoredSQL errors in multiple logs
20
© 2015 IBM Corporation
z Systems Software
New capabilities in 4Q, 2015
21
General capabilities (delivered via IBM Operations Analytics – Log Analysis and included with IOAz) Additional real-time alerting actions: SNMP Traps, EIF EventsRole-based access controlSupport for multiple time zones and time intervalsService Desk Extension: Incident and service request analytics
z/OS capabilities (included in the z/OS Insight Pack)Additional CICS insights from SMF 110 and EYULOGAdditional network insights from NetView netlogSecurity insightsPattern-based configuration for z/OS Log Forwarder job log data gathererAdditional out-of-the-box searches for DB2 and MQTranslation of z/OS Insight Packs (English + 10 languages) and documentation
© 2015 IBM Corporation
z Systems Software
Alerting actions: SNMP Traps, EIF Events
22
Benefit:– Utilize your existing event management tooling to track, highlight, enrich,
correlate and act upon conditions that are identified in their operational data by IBM Operations Analytics for z Systems through the use of SNMP Traps, Informs or EIF events.
– Broaden your scope of automation. Use NetView or other automation tools to take automatic action on any messages or other operational data as long as that data is consumed by IBM Operations Analytics for z Systems. This expands your current automation capabilities to automate on ANY data source that is fed into IBM Operations Analytics.
IOA now enables you to generate SNMP Traps and EIF Events.This is in addition to existing notifications (text, email, etc.)
© 2015 IBM Corporation
z Systems Software
23
Role-based access control and audit
Benefit:– Role-based access control and auditing capabilities enable customers to
maintain compliance with their data segregation and access control requirements. It is of special interest for service provider environments in which segregation of data is of particular importance.
© 2015 IBM Corporation
z Systems Software
Support for multiple time zones and time intervals
24
Benefit:– All users connected to a single IOA Log Analytics server, regardless of their
location, are able to view search results and graphs in their local time zone or in a different time zone of their choice. This new capability is particularly helpful for teams that are distributed across multiple time zones.
– Applications can now specify more than a single occurrence of a relative time interval. Instead of specifying “Last Day”, applications can specify “Last 3 Days” for example.
© 2015 IBM Corporation
z Systems SoftwareIntegration with Service Management Solutions
25
IOAz integrates with Monitoring, Automation and Event ManagementAutomation
NetView / SA (or other Automation tooling) – Receive and enrich, action or forward Events from ANY log source (not just Syslog)
Event Management Netcool Operations Insights (NOI)
– Launch to IOAz to analyze logs and metrics (IOA is included with NOI)– Search and analyze Events– Receive, correlate, enrich and action Events from IOAz (NOI or other Event Management
System)
Incident Management IBM Service Desk (or other incident management / trouble ticketing solutions)
– Generate Events to create Trouble Tickets– Analyze Trouble Tickets
Monitoring OMEGAMON
– Launch in context to IBM Operations Analytics from OMEGAMON and ITM workspaces– OMEGAMON Insight Pack to analyze ITM logs (RKLVLOG)
Service Management Unite (included with Performance Management and Service Management Suites)– Launch in context to analyze logs and SMF data in context of performance problem diagnosis
© 2015 IBM Corporation
z Systems Software
Using IOAz to broaden the scope of
Event Management and Automation
26
© 2015 IBM Corporation
z Systems Software
Event processing Generate Events from ANY log message(s) or other data in IOA Notifications can be in the form of:
– Text message– Email– SNMP Trap – EIF Event
Enhance your Visibility & Automation Capabilities
27
Increase scope of log monitoring and automation
Improve event correlation
IOAz can generate notifications for messages from any log in your enterprise.
Be Proactive!
© 2015 IBM Corporation
z Systems Software
28
Getting the most out of IOA notification capabilities
IBM Operations Analytics provides the ability to generate events based on messages, combination of messages over time, number of occurrences, etc.
Notifications can be generated from any data source:– Messages from Mainframe and Distributed Logs– SMF data– Events– Other
Examples include:– Send an email or text message whenever a specific message(s) is written
to a log … For example, message IRRB069I (RACF is being shut down) – Generate a SNMP Trap or EIF event when there are more then 500 failed
logon attempts in a 30 minute period
© 2015 IBM Corporation
z Systems Software
29
Notifications
Index alert action (ie send events back into IOA so they can be searched) You can use the index alert action template to index any triggered alerts. Email / Text alert action You can use the email template to send an email when a condition is met. Emails can easily be sent as text messages by most carriers
EIF alert action You can use the EIF template to send an EIF formatted event when a condition is met.
SNMP Trap alert action You can use the SNMP Trap template to send an SNMP Trap when a condition is met.
Script alert action You can use the Script template to execute a custom script when a condition is met.
Write to Log alert action You can use the Write to Log template to write an entry to a log file of your choice when a condition is met.
Send notifications in many forms …
© 2015 IBM Corporation
z Systems Software
Sending Events to any Event Receiver
Data Source 2
Alert Actions
IOA Server
Ingestion Pipeline
Alert Runtime
Index
Log
Script
SNMP
EIF
Event Management
System
(NOI, OMNIbusor other Event
Management tool)
Event Receiver
Automation
(NetView/SA or other
automation tool)
Data Source N
Data Source 1
30
IOA can generate standard SNMP Traps and/or EIF Events that can be
received and processed by ANY Event Receiver.
Event Receiver
Any Event ProcessorEvent
Receiver
© 2015 IBM Corporation
z Systems Software
31
Event Configuration is Simple
From IOAzSpecify the message or messages to trigger the NotificationChoose the event criteria (message IDs, number of occurrences, time period, etc.)Specify address of Event Receiver hostname/port or email address
From your automation toolCreate an automation statement(s) to:
– Enrich the Event– Forward the Event– Automate to correct the problem– Other
From your Event Management tool Enrich the Event Correlate with other Events and Log Messages Automate to correct the problem Create Trouble Tickets Other
Increase the scope of automation to include ANY log message
Most z/OS automation tools are limited to z/OS Syslog and Console messages
Correlate z/OS Events with Events from distributed systems to resolve problems
end-to-end
© 2015 IBM Corporation
z Systems SoftwareEvent driven automation scenarios
Optimized Performance
32
There are many scenarios where events can drive automation. Prior to IOAz, these scenarios were limited to events being driven from Syslog, because most z/OS automation tools only monitor the z/OS Syslog.
Since IOAz has access to many more logs than Syslog, we now have the ability to drive automation from messages coming from other logs and even other platforms.
We have included just a few examples in the subsequent slides. The possibilities are endless.
Benefit:• The subject matter expert can now access messages from ANY log in the enterprise• Events coming from IOA can be consumed by ANY Event receiver to automate, enrich,
correlate or forward Events or generate trouble tickets• Events can be generated in SNMP or EIF format. As a result, the events can be consumed
by any Event Receiver (Event Management or Automation tool). • Since IBM Operation Analytics for z systems can generate events from ANY message it
consumes and NetView can act as an event receiver, NetView can now automate on ANY log message (not just messages from Syslog). This scenario will work with any automation tool that can drive automation from events.
• This feature will enable customers the ability to ‘TAKE ACTION’ on any messages being consumed by IBM Operation Analytics for z Systems.
© 2015 IBM Corporation
z Systems Software
Alerting actions: SNMP Traps, EIF Events
33
Sample scenario for MQWebSphere MQ channel stopped abnormally
– MQ server runs on Windows with a MQ channel defined to MQ running on z/OS.– MQ server detects that the MQ channel to z/OS is not active and writes error
messages to the Windows MQ AMQError log. Subsequent MQ communications fail.
Without IBM Operations Analytics for z Systems:The ‘Channel down’ message is never proactively observed and the support team(s) struggle for hours to debug the problem and finally re-initiate the Channel.
With IBM Operations Analytics for z Systems:IBM Operations Analytics for z Systems detects the problem through MQ error messages written to the Windows MQ AMQError log.IBM Operations Analytics for z Systems generates an SNMP Trap (or EIF event) and forwards it to NetView (or other automation solution).Automation is driven from this event and resolves the problem by issuing a command to restart the MQ channel.
Outage avoided with IOAz!
© 2015 IBM Corporation
z Systems Software
Alerting actions: SNMP Traps, EIF Events
34
Sample scenario for DB2 DDFDB2 DDF applications timed out
– DB2 runs on z/OS; IBM Operations Analytics for z Systems collects DB2MSTR address space log.
– The customer applies bulk maintenance for z/OS and DB2 over the weekend. After application of maintenance, DB2 DDF applications experience time-outs.
Without IBM Operations Analytics for z Systems:Because maintenance occurs on a Saturday, operators do not catch the resulting problem until later. The DBA is notified on Saturday evening, a PMR is opened against IBM DB2, and diagnostics are started with the DB2 and TCP/IP L2 teams. By Monday morning, none of the agents can run transactions. DB2 and z/OS maintenance have to be backed out.
With IBM Operations Analytics for z Systems:IBM Operation Analytics is able to detect the time-out problem immediately after the maintenance is applied. Operators are notified immediately and are able to determine the root cause of the issue. End users do not experience downtime when they come into work on Monday morning.
Issue reported immediately with IOAz. Maintenance backed out. Problem avoided!
© 2015 IBM Corporation
z Systems Software
Search and analyze events, logs and metrics using IOA and Netcool Operations Insight.
Easily identify ‘related’ Events that may be candidates for suppression Identify “difficult to spot” seasonal events that often result in regular periodic
problems Easily identify which events occur in clusters Leverage visualizations that help you quickly isolate more sever and significant
problems.
Also provides opportunities for event reduction thus improving operational efficiency.
Log Analysis and Event Management in Netcool Operations Insight(IOA is included in the box with NOI)
35
© 2015 IBM Corporation
z Systems SoftwareLog Analysis – Streamline Incident Management
Incident ManagementThe traditional incident management process usually begins with one or more trouble tickets being opened for an incident (for example, slow response time for a specific application).
The first step is to engage the application support team and associated Subject Matter Experts for each of the application components (WebSphere, CICS, DB2, etc.). Each SME examine data from their specific subsystem and we usually experience a phenomenon commonly referred to as ‘ticket hopping’. During the ticket hopping phase, the trouble ticket will be reassigned multiple times before it lands on the correct SME’s lap. Over the lifetime of the incident, there is very little collaboration with respect to data and there’s usually a fair amount of ‘finger pointing’.
In the post mortem session, we usually conclude that the ‘time to resolution’ is very high and so is the number of people involved in the process of diagnosing the problem.
With IBM Operations Analytics for z Systems:IBM Operation Analytics will provide a unified view of the data, enabling the application support team to quickly focus on the problem component. The ability to search and analyze the data helps to quickly identify the problem area and the expert advice feature assists in finding the solution or workaround.
If an SME is needed for a specific component, you can transfer the ticket to that SME with the data that was surfaced by IOAz. Post mortem reveals that time to resolution is significantly decreased by as much as 50% with less involvement by the SME community.To be more proactive and improve mean time to recovery even more, the team can incorporate the use of IOA notifications to immediately notify (Text, Email, SNMP Trap or EIF Event) that a problem is occurring. Early detection will significantly decrease time to resolution and automation can be triggered to resolve the issue before the problem affects the end user.
© 2015 IBM Corporation
z Systems Software
Integration with Performance Monitoring
The One Two – Punch: Combine two very powerful tools to ensure performance and high availability of your enterprise.
Perform log analysis in context of OMEGAMON workspaces – This approach enables OMEGAMON users to perform in-context log analysis while doing problem determination
– From your OMEGAMON workspace, use the IOA search bar to search logs (using LPAR or Sysplex as the default context)
– Easy to implement - Configure TEP to display the IOA search bar
Launch IOA from OMEGAMON performance monitoring
workspaces to search logs in context
OMEGAMON + IBM Operations Analytics – Launch in Context from TEP
37
© 2015 IBM Corporation
z Systems SoftwareSearch and Analyze Operational Data in Context
38
Select a row first. In this example, a row
specifies a Queue Manager.
Specify a search string and timeframe
to analyze operational data from the
appropriate system(s)
© 2015 IBM Corporation
z Systems SoftwareAnalysis of Operational Data
39
IBM Operations Analytics analyzes log, metric and event data and surfaces insights
Built on industry expertise Expert Advice for faster time to resolution Expand analysis to include additional
data sources (from mainframe and distributed systems)
Launch into IBM Operations Analytics to
analyze logs and other operational data
to gain additional perspective and insights
and help diagnose root cause.
© 2015 IBM Corporation
z Systems SoftwareIntegration with existing Service Management solutions
(in a nutshell)
40
• Search and analyze logs, metrics, events and incident reports
• Launch from ITM, OMEGAMON, Service Management Unite & NOI
Surface anomalies
POWerful tools integrate to ensure performance and high availability of your enterprise.IBM zAware
Service Management Unite
Automation & Problem
DeterminationNetView
Performance Monitoring
ITM/OMEGAMON
Event ManagementOMNIbus/NOI
Incident ManagementControl Desk
Alert, enrich,
correlate and
automate
IBM Operations Analytics
© 2015 IBM Corporation
z Systems SoftwareSend us your logs!
Request a product demo using logs from your own test, development or production environments– IBM will load your logs into an IBM Operations Analytics server, then demo
the results back to you• A secure, dedicated drop box will be assigned to you• You will be sent detail upload instructions via email• Any file uploaded will be automatically moved to a dedicated IBM
Operations Analytics environment within 24 hours• All log data will be purged from the IBM Operations Analytics
environment within 48 hours after the demo event
To request your hosted demo, visit:http://services-useast.skytap.com:18280/WebDemo/
41
A guided demo is provided online at:
http://ibm.biz/ioazlivedemo
Or Take IOAz for a Test Drive
© 2015 IBM Corporation
z Systems Software
Announcing the IBM Operations Analytics for z Systems Early Access and Beta Program!
In 2015, we built on the strong foundation established over recent months as we develop and implement our product roadmap.
We are looking for customers and business partners worldwide who would like to help influence our roadmap and test new capabilities. The program is open-ended; interested participants may join at any time and stay on as long as they wish. That said, it is our desire to establish a set of “customer sponsor” relationships that will become instrumental in shaping the future of our offering.
To see the full program announcement, and to learn how to sign up, please visit us in our developerWorks community at:
https://ibm.biz/BdEkZV
IOA for z Systems Early Access and Beta Program
42
© 2015 IBM Corporation
z Systems Software
Additional IBM Operations Analytics Reference Material
Analytics Overview Video– https://www.youtube.com/watch?v=OQJapWiQECs
IOA for z Systems videos: – Overview: https://goo.gl/XsmGtq– Domain Insights: https://goo.gl/ktXYpL– Installation and Configuration: https://goo.gl/3QLm2N
IOA for z Systems Documentation– Knowledge Center:
• http://www.ibm.com/support/knowledgecenter/SS55JD
IOA – Log Analysis (server) Documentation– Service Management Connect
• http://www.ibm.com/developerworks/servicemanagement/ioa/log/index.html– Knowledge Center
• http://www.ibm.com/support/knowledgecenter/SSPFMY
43
© 2015 IBM Corporation
z Systems Software
44
© 2015 IBM Corporation
z Systems Software
More Detail and drill down to the next level
The remainder of the deck includes details for the following:
Bring your own DataFeeding additional logs into IOAOptionsExample using the Hardware Management Console log
IOAz 2.2Enhanced CICS insightsEnhanced Network insightsSecurity insightsLog Forwarder improvements
45
© 2015 IBM Corporation
z Systems Software
Bring Your Own Data
46
© 2015 IBM Corporation
z Systems Software
In addition to the out-of-the-box Insights for z/OS, DB2, IMS, CICS, MQ, Network and Security, IOAz can also be customized to meet the needs of your enterprise.
BYOD – Bring your own data– Enable Search and surface Insights for any text log (messages must have
time stamps)– Create your own Saved Searches – Create your own dashboards, graphs and charts– Gain access to any log data in the enterprise to debug end-to-end
applications, generate notifications and surface events
Enhance your Visibility – Avoid ‘Blind Spots’
47
Increase scope of log monitoring and automation
Remove ‘blind spots’
© 2015 IBM Corporation
z Systems Software
IBM Operations Analytics for z Systems IOAz provides a number of defined data sources to help you ingest
your data (out-of-the-box Insight Pack)– z/OS – Syslog (including CICS, IMS, DB2, MQ, Security, Network, etc.), CICS MSGUSR
and EYULOG, WebSphere sysout and sysprint, USS syslogd, SMF, NetView– Distributed Systems – MQ, DB2, Javacore, WebSphere, Service Desk, Microsoft SQL,
Active Directory, and many more. See the following URL for the latest list of Insight Packs available for IOA-LA: https://developer.ibm.com/itoa/resources/
But what if you want to add your own custom data types ? (BYOD)
• Custom Application Logs• Generated Report Files
• Statistical Records
• Others ?• Job Log data
48
© 2015 IBM Corporation
z Systems SoftwareExample: HMC Log as a data source
Customers often ask us if IOAz can consume logs from applications that they have written or even just some other log in the enterprise that is not currently supported by an existing Insight Pack.
The answer to this question is ‘YES!’. As long as it is a text log and the messages have a time stamp, they can be consumed by the IOA server.
Anyone can write an Insight Pack … And there are several options, depending on what you want to do with the data … In all cases, it is Quick and Easy.
We have received several requests to provide an Insight Pack for the HMC.
Retrieve the HMC logWe are using the HMC Log Tool (HLT) from the TechDocs library
– http://www-03.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/PRS5300– This tool allows you to retrieve log information using the zEnterprise Web Services APIs and
generates reports on the desired information.– Data is written to a file on USS (IOAz can consume USS files)
Feed the log data into IOAz for analysis and audit purposes.– Multiple options here depending on what you want to do with the data. All are simple.
49
© 2015 IBM Corporation
z Systems SoftwareWhat do you want to do with the data?
Search with Default Annotations (takes 10 minutes to configure)– Configure the z/OS Log Forwarder (data source type = ‘other’) to send the log data to IOAz.
The IOAz generic receiver will index and annotate the data to make it available for Search. – You can create and save quick searches, graphs and dashboards.
Annotate and analyze a CSV style log file (takes 20 minutes)– The DSV toolkit can be used to process any log file that follows a “Delimiter Separated
Value” format. Simply edit a text file describing the DSV layout.– Each separated value column becomes an annotated field in IOA– You can create and save quick searches, graphs and dashboards.
Annotate and analyze any text or log file (1-2 hours)– Allows for complete control on how files are read into IOA and parsed. This approach
allows for more sophisticated pattern searching and correlation. – Can be written in Java, AQL, Python, etc.– You can create and save quick searches, graphs and dashboards.
50
IOAz can consume and analyze any text log. There are 3 options to enable this depending on what you want to do with the log data
© 2015 IBM Corporation
z Systems Software
Example: HMC Insight Pack02/04/2015 14:54:13.300: 2007: User SooAcsadmin has acknowledged viewing l02/04/2015 15:31:59.860: 1100: The system clock has changed.02/05/2015 13:57:42.790: 734: Remote support call generated on USYS is bei02/05/2015 13:57:45.900: 674: Remote support call generated on USYS comple02/11/2015 14:08:01.680: 734: Remote support call generated on USYS is bei02/11/2015 14:08:02.810: 674: Remote support call generated on USYS comple02/11/2015 14:08:12.770: 734: Remote support call generated on USYS is bei02/11/2015 14:08:13.790: 674: Remote support call generated on USYS comple02/12/2015 09:14:18.810: 734: Remote support call generated on USYS is bei02/12/2015 09:14:39.470: 674: Remote support call generated on USYS comple02/12/2015 11:27:19.550: 1100: The system clock has changed.02/12/2015 12:32:16.650: 734: Remote support call generated on USYS is bei02/12/2015 12:32:19.350: 674: Remote support call generated on USYS comple02/12/2015 14:12:20.230: 722: An upgrade to EC level N98841 was performed.
Above is an excerpt from an HMC log. The log contains time stamps, event IDs and text.
In the following slides, we will show the difference between generic processing (Search with Default Annotations) and processing using the HMC Insight pack.
The HMC API was used to retrieve the log data and the logs were stored in the USS file system.
The z/OS Log Forwarder was configured to send the log to IOAz. 51
© 2015 IBM Corporation
z Systems Software
HMC Insight Pack – Search (using Generic Receiver)
52
Generic receiver discovered patterns in the logs.
Search capability generates helpful search results.
Graphical view of message volumes.
It take approximately 10 minutes to configure the log forwarder to send a log to IOA.
The generic receiver enables search and analysis, graphical views, save searches, expert advice, etc.
But in this case, it discovers too many generic patterns. We decided to write a quick insight pack to surface better insights.
© 2015 IBM Corporation
z Systems Software
Creating a Java based Insight Pack
Eclipse plugin available to assist in creating insight packs.
The plugin generates the required insight pack structure and supporting files.
All that is required is to write 2 pieces of java code and define the record structure of the resulting lines for IOA
– Splitter.java – Receives blocks of log data and breaks it up into distinct lines of data.
– Annotator.java – Receives each line of data and identifies the individual fields to be annotated for IOA.
Finally define what the name of the data source will be and map the splitter and annotator code using the supplied plugin.
Run an ant build using the provided ant build sample
53
It takes an approximately 1-2 hours to write an Insight Pack.
Time is significantly reduced after writing your first one.
© 2015 IBM Corporation
z Systems SoftwareExample: HMC Insight Pack in Java
02/04/2015 14:54:13.300: 2007: User SooAcsadmin has acknowledged viewing l02/04/2015 15:31:59.860: 1100: The system clock has changed.02/05/2015 13:57:42.790: 734: Remote support call generated on USYS is bei02/05/2015 13:57:45.900: 674: Remote support call generated on USYS comple02/11/2015 14:08:01.680: 734: Remote support call generated on USYS is bei02/11/2015 14:08:02.810: 674: Remote support call generated on USYS comple02/11/2015 14:08:12.770: 734: Remote support call generated on USYS is bei02/11/2015 14:08:13.790: 674: Remote support call generated on USYS comple02/12/2015 09:14:18.810: 734: Remote support call generated on USYS is bei02/12/2015 09:14:39.470: 674: Remote support call generated on USYS comple02/12/2015 11:27:19.550: 1100: The system clock has changed.02/12/2015 12:32:16.650: 734: Remote support call generated on USYS is bei02/12/2015 12:32:19.350: 674: Remote support call generated on USYS comple02/12/2015 14:12:20.230: 722: An upgrade to EC level N98841 was performed.
54
The 734 is the Event ID which would be useful to
annotate on
Annotated fields show up in the Search
Patterns
Fields are assigned to a key-value pair and can
show up in the grid view
An Insight Pack can surface important messages, text strings, text patterns, etc.
This is just a simple example.
© 2015 IBM Corporation
z Systems Software
Create custom quick searchesOnce the data is in IOA, it’s easy to create quick searches by simply running a search and saving it.
55
Save button lets you add any search to the saved
search menu
© 2015 IBM Corporation
z Systems Software
Creating custom graphs – Remote Support CallsTo create a graph of data, simply run a query, select column(s) to graph and then click the graph button and save it
56
The graph button allows you to graph 1 or more
data columns.
© 2015 IBM Corporation
z Systems SoftwareCreating custom dashboards
Multiple graphs can be saved to single dashboard
57
Distribution of Event IDs over time
Critical data backups
Login failures
Remote Support Calls
© 2015 IBM Corporation
z Systems SoftwareReference Material – Building Insight Packs
IBM Operations Analytics – Log Analysis resources web site– https://developer.ibm.com/itoa/resources/
An example of creating a Java based insight pack– https://developer.ibm.com/itoa/docs/log-analysis/insight-pack-creating
58
© 2015 IBM Corporation
z Systems Software
IOAz V2.2
More detail
59
© 2015 IBM Corporation
z Systems Software
Key enhancement: Enhanced CICS insights
60
© 2015 IBM Corporation
z Systems SoftwareEnhanced CICS Insights
61
IOAz consumes, analyses and provides insights using the following CICS data:CICS message data gathered from the following logs
– SYSLOG– MSGUSR– EYULOG
SMF 30 – CICS Jobs – CPU utilization– IO Rate– Paging Rate– Working Set
CICS SMF 110 data – CICS Monitoring Exceptions– CICS Global Transaction statistics (per APPLID)
In IOAz V2.2, the subject matter expert and/or application developer can gain insight and quickly isolate and diagnose CICS problems from a combination of CICS performance metrics and log data.CICS Monitoring Exceptions and Global Transaction Statistics are now available to aid in diagnosis.IOA notifications enable proactive monitoring for early detection and automation.
© 2015 IBM Corporation
z Systems SoftwareEnhanced CICS Insights
62
New Quick Searches for CICSIOAz V2.2 includes the following ‘Quick Searches’ (in addition to the current set of predefined searches for SMF30)Display all CICS ExceptionsDisplay CICS Wait on Storage ExceptionsDisplay CICS Policy ExceptionDisplay CICS Transaction Interval Summary Display CICS End of Day Summary- last weekDisplay CICS Transactions summaries where Tasks = MAXTASKS.
New CICS DashboardsCICS Regions Dashboard
– Interactive dashboard that integrates SMF and log data to diagnose:• Wait on Storage events per region over last day• Exceptions by Resource ID over last day• Short on Storage per region over last day• Tasks at Maximum Threshold over last day• Storage Violations per region over last day
CICS Regions Transaction Dashboard– Static dashboard showing:
• Transactions – top 5 regions over last day• Transactions – Max and Average per region over last day
© 2015 IBM Corporation
z Systems Software
CICS Demo Scenario
63
Eric the application developer is investigating an issue with his application running slowly.Prior to IOAz, Eric would have had to browse through each log looking for information on the state of CICS in general and his application specifically. This might involve the SYSLOG for the system and MSGUSR and EYULOG for each CICS region.If his application spanned multiple systems, he would have to log on to each system first and again manually browse each SYSYLOG and MSGUSR and EYULOG for each CICS region.He would be required to log on to yet a separate application to view SMF data.Once he has identified all the data sources, he will then need to manually correlate the data across the multiple sources to diagnose the problem.
© 2015 IBM Corporation
z Systems Software
CICS Demo Scenario, continued
64
With IOAz, Eric can logon to a single user interface and access all of the relevant logs in the enterprise and also reference SMF data to determine if an associated performance problem or CICS monitoring exception has occurred. First he checks the CICS Jobs dash board and does see a spike in CPU, I/O Rate and Working Set size. He looks at the Transaction dashboards and sees a corresponding spike in the transactions. He then looks at the Regions dashboard and does see some exceptions related to storage issues. Double clicking on the exceptions (insights automatically surfaced on the left side of the user interface) he can see most are related to the progname ICC$HELThen he checks the integrated IOA log search. This allows him to quickly search the SYSLOG, MSGUSR, EYULOG and SMF. In here he can search specifically for his application name to see where problems might lie, in this case ICC$HEL To be more proactive, Eric can configure a notification when a message, combination of messages or additional data indicates a problem. Notifications can be in the form of a text message, email, SNMP Trap or EIF Event and can be used to automatically resolve an issue, gather additional data or generate a trouble ticket.
© 2015 IBM Corporation
z Systems Software
Key enhancement: Enhanced network insights
65
© 2015 IBM Corporation
z Systems SoftwareEnhanced Network Insights
66
Network insights were first introduced in IOAz V2.1 to enable the search and analysis of z/OS network data provided by VTAM, TCPIP and syslogd message logs. In IOAz V2.2, you can now ingest, search and analyze NetView message data; specifically the NetView messages that are written to the NetView netlog. Customers can quickly and easily identify issues that are surfaced only through NetView netlog without actually going to NetView or searching the NetView Canzlog. In addition to network messages, the NetView Netlog contains other messages; for example, messages related to NetView automation failures.Since IOAz can consume logs from multiple systems, you can perform a single search across all logs to determine if a problem is occurring on more than one system.IOA notifications enable proactive monitoring for early detection of network or automation-related issues.
© 2015 IBM Corporation
z Systems SoftwareEnhanced Network Insights
67
New Quick Searches for NetView NetlogIOAz V2.2 includes the following ‘Quick Searches’ for NetView Netlog (in addition to the current set of predefined searches for Network insights provided in IOAz V2.1)All NetView for z/OS MessagesAll NetView for z/OS Action, Decision or Error messagesNetView for z/OS messages that indicate automation table violationsNetView for z/OS messages that indicate command authorization table violationsNetView for z/OS messages that indicate resource limits or storage thresholdsNetView for z/OS messages that indicate insufficient access authority or security environment violations
New NetView Netlog DashboardsNetView Message Counts - Top 5 per hour over Last Day NetView Message Type Counts - Top 5 per hour over Last DayTotal NetView Message Counts per hour over Last DayNetView Messages by Hostname - Top 5 per hour over Last Day NetView Message Types by Hostname - Top 5 over Last DayTotal NetView Messages by Hostname per hour over Last Day
© 2015 IBM Corporation
z Systems Software
Insights from NetView netlog
68
User Scenario 1Alice as the Subject Matter Novice is notified of a problem with the NetView automation table. A new automation statement has been added to the automation table to perform an automation action on an autotask when a specific message is processed. Even though the message is processed, the automation action did not occur.
Instead of signing on to NetView, Alice uses IBM Operations Analytics for z Systems and issues a Saved Search for NetView Automation to look for possible errors.
The search results find a NetView message: DWO032E AUTOMATION ACTION action COULD NOT BE ROUTED TO TASK(S) task.
Alice determines from this message that the problem is not with the actual automation table statement but instead is with the autotask where the automation action is to be performed. Next she can search for the specific autotask in the NetView messages to ensure the autotask is configured properly and started.
Alice can search and analyze ALL of the NetView logs at one time by issuing a single search command from IOAz.
Alice can create automation to check for these messages in the future and then issue the command to restart the autotask if it is not already started.
© 2015 IBM Corporation
z Systems Software
Insights from NetView netlog
69
User Scenario 2Jim as the Subject Matter Expert needs to determine if there were any attempts by NetView operators to issue unauthorized commands over the last week. Even though a NetView operator is allowed to sign on to NetView, an operator can be assigned access to a specific set of commands.
Instead of going to each NetView domain and looking for related security messages in the NetView Canzlog, Jim uses IBM Operations Analytics for z Systems and issues a Saved Search for NetView Security to look for unauthorized command attempts.
Jim can search and analyze ALL of the NetView logs at one time by issuing a single search command from IOAz.
The search results show Jim that there were multiple attempts to issue unauthorized commands on over the last week.
– BNH232E 'userid' IS NOT AUTHORIZED TO ISSUE COMMAND 'command’– BNH233E THE COMMAND 'command' IS PROTECTED BY COMMAND IDENTIFIER
'commandid' IN 'auth_method’– DSI213I ACCESS TO 'object' IS NOT AUTHORIZED
© 2015 IBM Corporation
z Systems Software
Insights from NetView netlog
70
Search results
No typing necessary. Just click on one of the IBM-provided quick searches or a saved
search
Analysis reveals patterns which can be combined to
build more complex searches; simply by selecting them.
Search logs from multiple systems or
choose to see results from specific systems
© 2015 IBM Corporation
z Systems Software
Key enhancement: Security insights
71
© 2015 IBM Corporation
z Systems SoftwareSecurity Insights
72
IOAz V2.2 introduces security insights so the subject matter expert, security administrator and/or application developer can quickly identify RACF failures and security issues. RACF failures often contribute to failed applications (for example, failure in read/write operations due to missing or incorrect RACF authorization).RACF security issues such as invalid authority and a significant number of invalid logon attempts can be a sign of a security breach.IOA notifications enable proactive monitoring for early detection of RACF security issues.
© 2015 IBM Corporation
z Systems SoftwareSecurity Insights
73
New Quick Searches for Security-related issuesDisplay all RACF MessagesDisplay all RACF Action Error or Warning messagesDisplay all Insufficient Access errors
– User attempted to access something but does not have the proper access authority Display all Insufficient Authority errors
– User attempted to execute a program but does not have the proper authority to do soDisplay all Invalid Password messages
– Any variation of the ICH/IRR messages dealing with invalid passwords.
New Security Dashboards Security Message Counts - Top 5 per hour over Last Day Security Message Type Counts - Top 5 per hour over Last Day Total Security Message Counts per hour over Last Day Security Messages by Hostname - Top 5 per hour over Last Day Security Message Types by Hostname - Top 5 over Last Day Total Security Messages by Hostname per hour over Last Day
© 2015 IBM Corporation
z Systems Software
Key enhancement: Pattern-based z/OS Log Forwarder data source configuration
74
© 2015 IBM Corporation
z Systems Software
Log collection configuration for complex environments ... made easy
75
Log Forwarder configuration has been enhanced to save time, be more dynamic and less error-prone.
The System programmer is responsible for configuration of the z/OS Log Forwarder(s). Depending on the number of log files being sent to the IOA server, the Log Forwarder configuration can be time-consuming. Today, it is a manual task and it can be error-prone.
Log Forwarder configuration has been enhanced in IOA V2.2 to support a wildcard and discovery feature to save time and decrease the possibility of configuration errors.
Significant reduction in time to configure log collection and analysis from large environments
© 2015 IBM Corporation
z Systems Software
Log collection configuration for complex environments ... made easy
76
Zach the System Programmer has a large number of WebSphere Application servers, or a large number of CICS regions, and he wants to collect data from all of them.Prior to IOAz, Zach would have had to configure the z/OS Log Forwarder for each and every job name.–This can be a time-consuming process, even when using the supplied Configuration Assistant. If you have 50 job names to configure, and each takes just two minutes, it will take an hour and 40 minutes to configure all 50.–This can be error-prone given the manual nature of the task – mistyping a job name means that job logs are not being ingested.–If a new WebSphere Application Server or CICS region is added after the z/OS Log Forwarder is started, the z/OS Log Forwarder must be restarted to pick up the new job log gatherer definition.
© 2015 IBM Corporation
z Systems Software
Log collection configuration for complex environments ... made easy
77
With IOAz, Log collection configuration is much easier. Zack can now create a single job log gatherer definition with one or more wildcard characters in the Job Name field. This definition serves as a template for all jobs that match the job name pattern.A handful of definitions can now cover many job names. Instead of an hour and 40 minutes to create definitions for 50 job names, it may take a minute or two.Valid wildcard characters are:
– * which represents any sequence of zero or more characters– ? which represents any single character
The z/OS Log Forwarder will start a data gatherer internally for each job name on JES spool that matches the wildcard job name value.The z/OS Log Forwarder continues searching for job names that match the pattern even after initialization has been completed. If a new job name appears it will attempt to start a job log gatherer for it.