Fast OLAP for Real User Measurement of Website Performance

• US 20140280880A1
• Filed: 03/14/2013
• Published: 09/18/2014
• Est. Priority Date: 03/14/2013
• Status: Active Grant

• Current Assignee: Akamai Technologies, Inc.
• Original Assignee: Soasta Incorporated (Akamai Technologies, Inc.)
• Inventors: Brewer, Vernon Archie III, Tellis, Philip Santosh
• Granted Patent: US 9,772,923 B2
• Time in Patent Office: Days
• US Class Current: 709/224
• CPC Class Codes: G06F 11/3082 the data filtering being achieved by aggregating or compressing the monitored data
• CPC Class Codes: G06F 11/32 with visual or acoustical indication of the functioning of the machine
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 11/3438 monitoring of user actions tracking the activity of the user H04L67/535
• CPC Class Codes: G06F 11/3452 Performance evaluation by statistical analysis
• CPC Class Codes: G06F 11/3495 for systems
• CPC Class Codes: G06F 2201/875 Monitoring of systems including the internet
• CPC Class Codes: G06F 2201/88 Monitoring involving counting
• CPC Class Codes: H04L 41/5009 Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
• CPC Class Codes: H04L 41/5083 wherein the managed service relates to web hosting

Abstract

A method for real-time capture and analysis of website performance metrics includes receiving, in each of a plurality of result servers, data values transmitted from a plurality of client devices, each data value representing a measurement of a performance metric associated with an experience of a user on a web page or using a web application, the measurement being made and the data value produced being transmitted in real-time while the user is on the web page or using the web application. For each data value received, a count in one of a set of N data buckets defined in a memory is incremented. All of the counts in each of the corresponding data buckets is aggregated to produce a histogram of the performance metric is rendered on an analytic dashboard.

Claims

1. A method for real-time capture and analysis of website performance metrics comprising:

receiving, in each of a plurality of result servers, data values transmitted from a plurality of client devices, each data value representing a measurement of a performance metric associated with an experience of a user on a web page, or using a web application or mobile app, the measurement being made and the data value produced being transmitted in real-time while the user is on the web page, or using the web application or mobile app;

for each data value received, incrementing a count in one of a set of N, where N is an integer greater than 1, data buckets, each data bucket being defined in a memory for data values falling within a predetermined range of values, the set of N data buckets being arranged in an order of data value ranges;

transmitting, by each result server, the count of each of the N data buckets to a unit over a network;

aggregating, by the unit in a final set of N data buckets, all of the counts received from each results server for each corresponding data bucket such that an ith data bucket of the final set contains a count representing the sum of all of the ith data buckets received from all of the results servers; and

rendering, in real-time on an analytic dashboard, a histogram of the performance metric, wherein each bar of the histogram corresponds to one or more of the N data buckets.

2. The method of claim 1 further comprising the automated steps of:

determining a total count equal to the sum of the counts of all N data buckets in the final set;

calculating a percentile count as the product of the total count and a percentage M;

identifying a particular data bucket of the final set as having a Mth percentile data value based on the percentile count;

approximating the Mth percentile data value as;

V_B+((R/NUM)×

I)

where V_B is a value at a bottom of the predetermined range of the particular data bucket, R is the predetermined range of the particular data bucket, NUM is the count of the particular data bucket and I is an incremental count within the particular data bucket that corresponds to the percentile count.

3. The method of claim 1 wherein the set of N data buckets and the final set of N data buckets are identically configured.

4. The method of claim 1 wherein the performance metric comprises web page load time.

5. The method of claim 1 wherein the predetermined range is identical for all of the N data buckets.

6. The method of claim 1 wherein at least one of the data buckets has a different predetermined range than that of another data bucket.

7. The method of claim 1 further comprising scaling a height of each bar of the histogram such that the area of a given bar is equal to the number of data points represented by the given bar.

8. The method of claim 1 further comprising:

selecting a minimum bar width for the histogram; and

in the event that a given bar of the histogram has a width that is an integer multiple of the minimum bar width, scaling a height of the given bar to a reduced height equal to the number of data points represented by the given bar divided by the integer multiple.

9. The method of claim 8 further comprising storing the final set of N data buckets in a database.

10. The method of claim 9 wherein the database is accessible to a computer running a program that generates the analytic dashboard.

11. The method of claim 1 wherein the order is an ascending order.