Mainframe performance analysts spend most of their time ensuring that their companies’ business-critical applications run smoothly, without delays or disruptions. Even under normal circumstances this can be difficult.
But under peak periods or times with higher workload stress, the task may feel too insurmountable to overcome. Some analysts may even think that application availability issues during the busiest periods (like Cyber Monday) are inevitable.
Just wrap your head around these 2018 statistics from Adobe Analytics:
- Cyber Monday brought in $7.9 billion (making it the biggest online shopping day in the US)
- Black Friday brought in $6.2 billion (23.6 percent growth YoY) in revenue
- The three hours between 10:00 p.m. and 1:00 a.m. [Eastern Time Zone] on Cyber Monday are expected to drive $1.7 billion in online sales, roughly $300 million more than an average full day during the year.
- Sales coming from smartphones hit an all-time high of $2 billion on Cyber Monday
“The three hours between 10:00 p.m. and 1:00 a.m. ET on Cyber Monday are expected to drive $1.7 billion in online sales”
What was once a single day – Black Friday – of massive retail sales, has now turned into an entire Cyber Week (starting the Monday before Thanksgiving, through Cyber Monday), where every day is likely the busiest shopping day of the year, in succession.
The question is, is your infrastructure ready?
Cost of IT Downtime
The profit potential for retailers during this period is massive, however, the potential profit loss for any period of application unavailability is also massive.
According to ITIC, who has been surveying businesses yearly since 2008, a single hour of downtime costs at least:
- $100,000 for 98% of businesses
- $300,000 or higher for 86% of businesses
- $1 million (or more) for 34% of businesses
(This data comes from ITIC’s “2019 Global Server Hardware, Server OS Reliability Survey”, which polled over 1,000 businesses worldwide from November 2018 through January 2019, and does not include any litigation, fines, or civil/criminal penalties that may arise from lawsuits or non-compliance issues.)
These numbers are dramatic by themselves, but when considering that they’re the average for the year, it only emphasizes the importance of ensuring that your infrastructure can handle the workload peaks during Cyber Week, where every minute can cost millions in lost revenue.
What You Can Do to Ensure Application Availability Throughout the Year
In his latest blog, Jerry Street pointed out that there’s much more to ensuring a healthy environment than just making sure you have enough CPU.
Many businesses treat the week (or so) before Black Friday as a sort of “fire drill,” but that’s often too late to ensure your mainframe infrastructure is ready. It’s important to start working on this today – not when you start seeing the workloads increase.
Here are some useful tips you can use:
Know your sales vs. shopping workloads / Refine your workloads
If you have a shopping transaction vs. a sales transaction, you may want to refine the service classes to ensure you give the sales transactions more priority than shopping. Refining your workloads into appropriate service classes will minimize the impact to sales when resource limitations are hit by surprisingly larger demands.
Ensure you have enough headroom
Compare key metrics for an average day before your last peak period to the highest day of your last peak period. Take that delta and apply it to your current average day leading into your next peak period. It’s a given that workloads grow over time. Make sure that you have at least enough headroom to absorb your last peak period. More preferable would be enough headroom to absorb your last peak period plus 10% or 15%. That way you can accommodate if things go better than expected.
Create dashboards for key metrics
Create dashboards containing key metrics for your peak period. This will allow other teams, such as your NOC, the ability to monitor them, freeing you up to do other things in preparation for your peak period. More general “roll-up” dashboards should also be shared with management, providing them with peace of mind.
Remember that severely poor performance can have as much of an impact as an outage.
Let AIOps solutions help
Take advantage of AIOps solutions such as IntelliMagic Vision for z/OS that automatically highlight potential issues before they impact availability and help you resolve them quickly and efficiently.
Prepare for year-round peaks
Cyber Monday may be the largest shopping day in the US, but there are still known (and unknown) peaks that impact other industries throughout the year. These peaks not only include higher risk to your business but can greatly impact your ability to enjoy the holidays stress free without being called in for an emergency disruption.
To learn more on this topic, and to ensure that your infrastructure is prepared for these peak periods, watch this webinar for Top z/OS strategies for Peak Workloads like Black Friday and Cyber Monday.
Subscribe to our Blogs
Subscribe to receive monthly updates about the latest industry news and high quality content, like webinars, blogs, white papers, and more.
Health Insights and Change Detection
To help enhance the value you can derive from that SMF data, IntelliMagic Vision provides two complementary types of automated anomaly detection: health insights and change detection.
CICS Max Tasks: Automated Change Detection Use Case 7
Change Detection and Health Assessments (based on pre-defined best practice conditions) play complementary roles, but there can be scenarios where they both flag issues. This use case with CICS Max Tasks is one such example.
Batch Job Elapsed Time: Automated Change Detection Use Case 6
Elapsed and CPU times can be of interest for batch jobs, especially those on the critical path for key business job cycles. This use case reflects an abnormally long run time for a daily batch job, and illustrates how potential causes might be investigated.