Are you interested in what happens when the Christmas rush ends? That’s exactly right, we start receiving information about the holiday’s impact not only on our wallets, but also on any kind of work process. This article is no exception. The illustrations presented are made based on December data and the Monitis users’ data library. Since Monitis specializes in cloud based monitoring products, the target parameter is response time. That’s the reason the analysis will be implemented from website response time records.
Do we need an answer to the query provided? The quicker and clearer the answer, the better the orientation. Below are presented several tables showing various statistical metrics from Monitis which can be helpful in figuring out December’s Christmas rush pattern. I would like to mention: the lower the response time, the better the performance.
The charts and tables speak for themselves, and loudly!
a) To monitor website response time records, we need to observe the percentage changes specifically, on a daily basis; a small change occurred on the 3rd of December, nevertheless response time records were quickly restored. Notice that all the other values fluctuate slightly which means that response time, overall, is constant. See the next chart.
To present the response time fluctuations, we averaged the response time records on a daily basis. So, each point on the graph represents the average response time on the day indicated. The Monitis users’ data library serves as the source for the analysis. The blue line indicates the average response time for the entire month. Therefore, we can compare the daily average response times with the monthly one. Furthermore, the chart illustrates some degree of regularity. Looking through it thoroughly, we find weekly fluctuations, i.e. on weekends the response times show really good performance due to the lower number of user visits and modest WebPage load volumes. On the other hand, when the holidays come to the fore, the response times observed are lower than the expected value.
b) Are we interested in percentiles? The parameter describes response time thresholds, i.e. 50% of response time records are below 396 milliseconds. In our case, 25th 50thand 75th percentiles are considered.
But what about on a country basis?
The pie chart below represents webpage test volume proportions of Monitis users. The greatest share, 56%, belongs to the U.S.
The global average response time is associated with different countries’ average values. Notice that the longest reaction time is in Brazil.
Summarizing all the data, we can see that December, and particularly Christmas time, does not have a huge influence on the monitoring patterns. On average, the daily response times change in value is 0%. This illustrates the reliability of the monitoring features and patterns.