Tag Archives: Oracle
Rittman Mead Sponsors Georgia Oracle Users Group’s Tech Day
As part of Rittman Mead’s continuing support and participation in the Oracle community, we are proudly sponsoring this year’s Georgia Oracle Users Group’s Tech Day on Thursday, March 3, 2016, in Downtown Atlanta at the Executive Conference Center.
Rittman Mead’s own Consulting Manager, Andy Rocha, will be speaking at this year’s event. Andy will be presenting on how we created a game-tracking page in OBIEE for the 2014 World Cup. He will break down how we created the dashboard to track real-time scores, stats, match events, and player and team information.
We will also have a booth set up, so if you are in the area, come down to see us and get a demo of some of our exciting new products and services.
To learn more about GaOUG and their events, visit their website here.
The post Rittman Mead Sponsors Georgia Oracle Users Group’s Tech Day appeared first on Rittman Mead Consulting.
OBIEE Performance – Why Metrics Matter (and…Announcing obi-metrics-agent v2!)
One of the first steps to improve OBIEE performance is to determine why it is slow. That may sound obvious—can’t fix it if you don’t know what you’re fixing, right? Unfortunately, the “Drunk Man anti-method”, in which we merrily stumble from one change to another, maybe breaking things along the way and certainly having a headache at the end of it, is far too prevalent. This comes about partly through unawareness of a better method to follow, and partly encouraged by tuning documents comprising reams of configuration settings to “tune” and fiddle with without really knowing why or how to prove if they indeed actually fixed anything…
Determining the cause of performance problems is often a case of working out what it’s not just as much as what it is. This is for two important reasons. Firstly, we begin to narrow down the area of focus and analysis. Secondly, we know what to leave alone. If we can prove that, for example, the database is running the query behind a report quickly, then there is no point “tuning” the database, because the problem doesn’t lie there. Similarly, if we can see that a report taking 60 seconds in total to run spends 59 seconds of that in the database, fiddling with Java Heap Size settings on OBIEE is going to at the very, very most reduce our total runtime to…59 seconds! This kind of time profiling is important to do, and something that we produce automatically in our Performance Analytics Report:
So, how do we pinpoint what is, or isn’t, going wrong? We need data, and specifically, we need metrics. We need log files, too, maybe for the real nitty-gritty of explain plans, but a huge amount can be understood about a system by looking at the metrics available.
Any modern operating system, from Windows to Linux, AIX to Solaris, will have copious utilities that will expose important metrics such as CPU usage, disk throughout, and so on. These can often be of great assistance in diagnosing performance problems.
OBIEE DMS Metrics
When it comes to OBIEE itself, we are spoilt by the performance counters available that since 11g (and still in 12c) have been exposed through the Dynamic Monitoring System (DMS). They were even there in 10g too, but accessed through JMX. These metrics give us information ranging from things like the number of logged in users, through how many connections are open to a given database, down to real low-level internals like how many threads are in use for handling LDAP lookups. Crucially, there are also metrics showing current and peak levels of queueing within the various internal systems in OBIEE, which is where DMS becomes particularly important.
By being able to prove that OBIEE has, for example, run out of available connections to the database, we can confidently state that by changing a given configuration parameter we will alleviate a bottleneck. Not only that, but we can monitor and determine how many connections we really do need at a given workload level. The chart below illustrates this. The capacity of the connection pool is plotted against the number of busy connections. As the number of active sessions increases so does the pressure on the connection pool, until it hits capacity at which point queueing starts—which now means queries are waiting for a connection to the database before they can even begin to execute (and it’s at this point we’d expect to see response times suffer).
So this is the kind of valuable information that is just not available anywhere other than the DMS metrics, and you can see from the above illustration just how useful it is. To access DMS metrics in OBIEE 11g and 12c, you have several options available out of the box:
- DMS Spy Servlet
- This includes the very useful (but undocumented) option to pull the metrics out in XML format, by including
format=xml
as a request parameters—thanks to etcSudoers on the #obihackers IRC channel for this gem!
- This includes the very useful (but undocumented) option to pull the metrics out in XML format, by including
- Fusion Middleware Control
- WLST
- opmnctl (not 12c)
- There’s also the paid-for option of the BI Management Pack.
Some of these are useful for programmatically scraping the data, others for interactively checking values at a point in time.
obi-metrics-agent – v2
At Rittman Mead, we always recommend collecting and storing DMS metrics (alongside others, including OS) all the time—not just if you find yourself with performance problems. That way you can compare before and after states, you can track historical trends—and you’re all set to hit the ground running with your diagnostics when (if) you do hit performance problems.
You can capture DMS metrics with the BI Management Pack in Enterprise Manager, you can write something yourself, or you can take advantage of an open-source tool from Rittman Mead, obi-metrics-agent.
I wrote about obi-metrics-agent originally when we first open-sourced it almost two years ago. The principle in version 2 is still the same, we’ve just rewritten it in Jython so as to remove the need for any dependencies like Python and associated libraries. We’ve also added native InfluxDB output, as well as retained the option to send data in the original carbon/graphite protocol.
You can run obi-metrics-agent and just write the DMS data to CSV, but our recommendation is always to persist it straight to a time series data store such as InfluxDB. Once you’ve collected the data you can analyse and monitor it with several tools, our favourite being Grafana (read more about this here).
As part of our Performance Analytics Service we’ve built a set of Performance Analytics Dashboards, making available a full-stack view of OBIEE metrics (including DMS, OS, and even Oracle ASH data), as seen in this video here (click on the image to enlarge it):
If you’d like to find out more about these and the Performance Analytics service offered by Rittman Mead, please get in touch. You can download obi-metrics-agent itself freely from our github repository.
The post OBIEE Performance – Why Metrics Matter (and…Announcing obi-metrics-agent v2!) appeared first on Rittman Mead Consulting.
Introduction to Oracle R Training
Rittman Mead is thrilled to announce our new Introduction to Oracle R courses!
With the launch of 12c comes an exciting extension to OBIEE’s analytical capabilities—embedded R execution. Oracle has made several investments in the R language and its related technologies (Oracle R, ROracle, ORAAH, and Oracle R Enterprise). Based on our experience deploying advanced analytics solutions from within business intelligence departments, we view this simplified pairing of OBIEE and R as a great way forward in delivering advanced analytics across organizations.
But the world of R is a big one, with hundreds of open-source R packages and countless training avenues. With so many confusing options on the market, we wanted to simplify things for you. It can be difficult to know where to start, so we’ve designed an R course to help you bridge the gap between existing BI/DW skills and the new skills required to confidently derive insights with R. We want to help you look beyond the dashboard and delve into novel analytical techniques.
The Advanced Analytics course dives deeply into the analytical capabilities of 12c and R. We’ve designed the training so that prior knowledge of R isn’t assumed. The course builds on existing SQL and data visualization skills—covering R use cases from predictive analytics and time series forecasting to natural language processing and many others.
We want to share with you our vision of R-enhanced business intelligence that provides more insights, predictions, and actionable discoveries, all while scaling effortlessly with your data. With our training, you’ll learn to expand your analytic skills and become an invaluable asset to your company’s BI operation.
We are excited about what R brings to the world of BI and can’t wait to share our knowledge.
We will follow up this blog post with one about the business value of our new Introduction to Oracle R course and another post about the more technical aspects of the course.
The post Introduction to Oracle R Training appeared first on Rittman Mead Consulting.
How to know the part numbers for OVS and OVM 3.2.9 installation?
Question: How to know the part numbers for OVS and OVM 3.2.9 installation?
Answer:
Oracle VM Server for x86 3.2.9 server ISO, search patch ID 16410428; and also available from ULN - https://linux.oracle.com
Oracle VM Manager 3.2.9 upgrade, search patch ID 16410417
Oracle VM Agent 3.2.9 for SPARC, search patch ID 16694949
Oracle VM Manager 3.2.9 Part number is V74420-01
Implementing OBIEE Commentary
My Application Development Experience with Rittman Mead
Greetings, readers! My name is Nick Padgett, and this is my first blog post with Rittman Mead. I started with Rittman Mead as an intern over a year ago while I was attending Kennesaw State, a local university, but I’ve graduated from both school and the internship. Since then I’ve been working on several interesting applications. Today, I would like to take some time to discuss developing one such application, our OBIEE Commentary tool, as well as my learning experiences working at Rittman Mead.
When I first started with Rittman Mead in October 2014, I was beginning my final year in a Computer Science program. When I joined, I knew very little about Business Intelligence but was eager to learn everything I could. Learning BI while still in school created a powerful synergy between my class studies and the real world. As I learned web development and security in class, I learned about data warehouses and agile development at work. In addition, I had the opportunity to practice my new knowledge and skills on a daily basis.
The most important lesson I learned during my time as an intern was the importance of developing primarily to meet the user requirements. This subject was often ignored in school, except for the obligatory Software Engineering course. But now, as a formal developer, I was always being reminded of the users we were developing for. All features for all projects were first qualified by asking “Is this what the user wants or needs?” or “Will this add value for the users?”
The first project I was assigned was the Rittman Mead commentary tool. As I assisted in development, the team always looked back to the initial requirements: “Users want commentary in OBIEE.” All development was focused in that direction, and any feature which hindered the goal, despite the technical impressiveness, was removed from the application.
Now, after spending many hours developing the commentary tool, the team is confident it has effectively answered the primary user requirements. We’re all very excited to provide this tool to our customers, and we are confident it will meet all the requirements we spent so long accumulating and developing for.
The Primary Use-case: OBIEE Commentary
In a typical environment, users are required to leave the application to communicate with their peers. This often involves screenshots, long text descriptions of the subject, and a service such as email or Slack. However, forcing users to leave the application in order to do their job is less than optimum. What’s to keep them using the expensive tool you purchased for them if they have to leave it every few minutes?
Many organizations recognize this issue and seek to implement commentary themselves. Frequent approaches typically use either a text object on a dashboard, a write-back enabled input, or even strange iFrame solutions to achieve the desired capabilities. Obviously, these implementations have serious drawbacks, and may not be maintainable or feasible for many organizations.
The commentary tool we have developed uses none of the above approaches, but instead uses a standard web deployment, hosted in WebLogic, to provide similar capabilities. This means no RPD modifications, no requirement for dashboard designers to become involved, and no security threats with iFrames or Cross-Site Scripting (XSS) attacks, while still fulfilling the primary use case of adding commentary to your OBIEE dashboards.
Our tool allows for users to create several different types of comments, all displayed through standard HTML in a web browser. Comments can be added on a dashboard to provide commentary for the page as a whole, while other comments can be placed on reports, which will be shown anywhere the report is located. You can even submit replies to existing comments, allowing for comment threads over a particular subject. All of these comments are available for use without having to leave the application.
A Second Use-case: Documentation
Some comments are highly structured and curated, containing a wealth of information. Some users like to add comments explaining how a measure is calculated, or a post-mortem for the launch of a new product. These comments are best classified as documentation, as they don’t serve to engage in a conversation, but to present information. In addition, there may be formal documentation that is required to be on a dashboard, and relates directly to a series of reports. The most common way to provide this is to supply links to external applications specifically suited for documentation. Among a myriad of issues pertaining to accessibility concerns, this approach forces users to leave the application, some of whom may never return.
As an alternative, documentation can be placed directly on a dashboard or report, but this leads to design problems. Having a twenty-page document on a dashboard is hardly an acceptable solution, even though this is the best place to put your documentation. In addition, users will have to write their documentation in HTML, or at least be able to understand it.
The commentary tool also allows for this use case. Documentation can be created or edited using a WYSIWYG (What You See Is What You Get) editor, similar to any standard document editor, and added to a “Table of Contents” on a dashboard. The Table of Contents lists all current documents available to view and displays them in a dedicated area. This provides access to any pertinent content directly on a dashboard, rather than forcing users to migrate to an external application, which is the entire point for having native commentary in the first place.
A Third Use-case: Integration
A common concern may be the introduction of yet another channel for communication. For example, many organizations use Slack as a communication tool or Atlassian Confluence for documentation. Rittman Mead does not desire to replace these tools or make them obsolete to your organization. We acknowledge the fact you spend money on these applications, and they may be widely adopted within your organization.
Instead of forcing you to maintain several channels of communication, our commentary tool allows integrations with applications such as JIRA and Confluence. Comments can be submitted as JIRA issues, and documents can be synchronized with Confluence pages. Several other integrations are available by default, while additional implementations are available on a per-customer basis. Now, if users absolutely must leave the application, they can do it on their own terms, with the convenience of a few clicks.
As I have grown during my time here with Rittman Mead, so has our commentary application. While I learned how to cater to user requirements, the tool evolved into an extension of this knowledge and was developed with the intention to meet all customer needs. I am incredibly excited to see this product released soon, and I can’t wait for users to start getting more out of their BI applications.
More information on this tool will be coming very soon, so keep an eye on our blog and website for updates!
The post Implementing OBIEE Commentary appeared first on Rittman Mead Consulting.