IT estates need to be optimised to run eﬃciently day-to-day and not to waste IT capacity. They also need to be analysed through their working cycles to eﬀectively understand how to avoid any capacity issues, particularly those leading to downtime or degradation of services. This is important for meeting SLAs, and regulatory requirements, and ensuring margin is protected.
The Application Support team in an Investment Bank were under pressure to run their IT estates more efficiently to minimise costs, while better managing capacity to prevent any failures or degradation of services. Not only was this about managing their systems’ capacity in real-time but also future proofing their business against potential adverse impacts.
Following poor performance of certain applications during a recent spike in trading activity, the Application Support team were concerned about their ability to not only see how their IT estates were performing in real-time but, more importantly, their ability to plan against future scenarios. In addition, the growth in virtualisation and the increased use of cloud based overflow means that the IT estate is variable in capacity and highly inter-related.
Introducing ITRS Insights
ITRS Insights performs an impressive range of analytics across both real-time and historical data at a scale required by today’s big data world. This makes it suitable for running analytics on your business activities, your underlying technology or a combination of both.
ITRS Insights can spot when things aren’t “normal” in a constantly changing environment. The rapidly changing variables involved mean that it is not enough to just set parameters as they aren’t able to respond dynamically to the changing environment. The valuable insights produced can answer key questions such as “will my IT estate cope with a large increase in the trading activity?”
In this scenario, ITRS Insights allows senior Application Support staﬀ to analyse the data in Insights to understand how diﬀerent variables interact with each other in real-time. These variables could include trades per second going through the estate, latency of trading, IT machine data such as CPU load, disk I/O, memory usage etc. Using mathematical correlation, it is possible to understand the interaction of these variables, and using prediction algorithms, to predict the volume of business activity which would generate a maximum threshold, either physical (100% CPU) or business (maximum allowable latency).
Real-time analysis of business and machine data;
- Variable correlation: ITRS Insights provides the algorithms to correlate any variables in real-time. The correlation may be linear in the early part of the loading of a system (CPU increases as trade volumes increase) but may deviate as the systems approach limits (IT systems are often non-linear in their performance characteristics). ITRS Insights is integrated with their existing ITRS Geneos estate, providing alerts if the IT systems deviate from expected correlations.
- Predictions: ITRS Insights supports the ability to predict the value of one variable which will achieve the value of another variable. In this scenario, the Application Support team may want to know the volume of trades per second which will generate a given limit of latency, or a given limit of CPU utilisation (e.g. 90%). ITRS Geneos has a rule set to detect the growth in both the measured and the predicted variable.
Secure and scalable;
- Immutable: your data streamed into Insights is stored in a read-only format, making it secure and immutable. Stored results of analysis are also read-only, adding to conﬁdence and security.
- Clustered architecture: the underlying architecture of ITRS Insights is a cluster of nodes. Both the data and processing are spread across the nodes, with no data or processing reliant on a single node only. If a node goes down then the cluster can continue without issue, ensuring constant availability of both data access and analysis.
Collaborative sharing of information: your queries, research and investigations are pulled together in one or a series of notebooks that can be easily shared with diﬀerent audiences and stakeholders, helping cross-team and cross-geography collaboration through sharing vital information. If the query is dynamic, then the visualisation will also be. This is particularly useful when sharing analysis and insights with a business focused audience, who might not necessarily have direct access to the ITRS Insights application.
Simple integration: ITRS Insights’ RESTful interface allows you to easily deploy and seamlessly integrate with a wide variety of data collection or visualisation tools, or your own proprietary systems. It also includes a seamless integration to your Geneos deployments.
Designed for a great knowledge experience: the easy to use interface is designed to enable fast and simple visibility into your complex environments. Making sophisticated analysis and deep insight, simple to access. ITRS Insights uses “domains” to understand the subset of data you want to work with. Easily selected “contributors and streams” let you select the sources of data that are important to you.
ITRS Insights is able to analyse business and machine data and to calculate correlations in real-time. From this, the Support Analyst can either look for breaks in expected correlations, or use the prediction capabilities to predict when the IT estate is likely to reach unacceptable levels of performance. Geneos can then be used to watch for these scenarios.
The analysis produced by ITRS Insights supports capacity planning by combining business volume data (trades per second) with load on the system (CPU utilisation) or latency of trading, and all of this analysed over time and compared with comparable trading days. Using this insight, the IT estate can be optimised to prevent wastage of expensive IT capacity, and equally importantly, ensure there is head room available to cater for peak load situations. Though they only occur occasionally, they must not cause failure or unacceptable performance degradation that could negatively impact the business as a whole.