The Session Overview page of the BMC Application Performance Management Console enables you to view details for a single session so that you can locate pages that have availability and performance issues and drill down to investigate the problems such as the following:
- A performance issue causes a page to fail one of the service-level thresholds (SLTs).
- An availability issue causes a page to fail to open or causes an error (for example, one of the objects on the page does not load).
The Session Overview page consists of the following sections:
Session Overview section
As shown in the following figure, the Session Overview section consists of the session summary section and the session timeline graph:
Session Overview section of the Session Overview page
The Session Overview section shows the following information about the session:
- User and group ID—String (for example, John Doe @ example.com) containing the user ID (John Doe) and group ID (example.com), which enables you to associate a session with a specific user. For more information, see User and group identification.
- Browser—Browser that the client uses. To see the version of the browser, mouse over the browser icon.
- Platform—Device platform that the client uses (desktop or mobile)
- Location—City and country where the user is located. You can also see the country flag and user location on the map:
- If the session has no issues, the dot on the map is colored green.
- If the session has an availability or performance issue, the dot on the map is colored red.
- Pages with Errors—Quantity and ratio of pages with availability issues in relation to all pages in the session
- Slow Pages—Quantity and ratio of pages with performance issues in relation to all pages in the session
Duration—Length of the session. This section also shows the status of the session, In Progress or Completed.
- Start—Time and date when the session started
- Throughput—Session throughput, in kilobits per second.
You can export the information about the session and save it to your computer for later analysis.
For a quick view of the session to locate problematic pages, you can use the session timeline graph, which displays all pages in the session chronologically. Sections on the graph are highlighted as follows:
- Blue—Pages without issues
- Red—Pages with performance or availability issues
- Gray—Interval with no opened pages
For the sessions that are still in progress you can see the pages that the user opened so far. After you refresh the page in the browser, in addition to the pages in the timeline, the graph also shows new pages that the user opened.
The color on the timeline is deeper when the density of pages in a given time period is high. High page density can cause difficulty in understanding what happened in a short interval. The timeline enables you to zoom in to identify problematic pages for investigation. To zoom in and out, hover over the session timeline and use the mouse wheel. When you zoom in, you can use the mouse to move left and right in the timeline.
To see the name for each page in the session, hover over the triangle marking the page, as shown in the following figure. The tooltip displays the page name and the type of problem.
Page with an availability issue on the timeline
Page Overview section
The Page Overview section consists of the page summary section and the object waterfall section.
The page summary helps you to understand and analyze the page design in relation to the problem:
Page summary section of the Session Overview page
If the HTTPS protocol was used for secure transmission of the page, the lock icon appears by the page name.
The latency doughnut chart helps you to identify how latency contributed to application performance problems. This chart shows the ratio of the following types of latency:
- Host—Time for the application to process the page
- Network—Time for the network to transmit the page. Use this latency to investigate the connection speed.
- SSL—Time for the web system to negotiate the SSL encryption for the page (applicable only if the page was encrypted)
- Idle—Period of inactivity. Use this latency to investigate periods of inactivity when the page loaded.
- E2E—Time for the request to be transmitted across a network from source to destination (end-to-end latency); for a page, the period from the user clicking to open the page to the last object loading on the page
PRT—Page Render Time (PRT); time for the browser to load the page, from loading the first object on the page to the last one. Use this latency to investigate why a page took a very long time to render, or what slows the loading of a page.
- ADT—Akamai Download Time (ADT) shows the time it takes to load the web content of a page coming from Akamai.
- Redirect—Time to redirect the client to this page. Use this latency to investigate the number of redirects before the application loaded the page.
The page summary also contains the following information:
- Service Level Threshold (SLT)—Identifies whether the performance is compliant with the service level. For more information about SLTs, see Page service-level thresholds. You can see the defined SLT level and whether it was exceeded.
- Objects with Errors—Shows the quantity and ratio of objects with errors (marked with red) in relation to all objects on the page
- Duration—Time taken to load the page (equal to E2E latency)
- Start—Time and date when the page started loading
- Throughput—Rate of loading the objects in the page over the communication channel, in kilobits per second
- Size—Page size, in kilobytes
When you select a page in the session timeline graph, you can view all objects that the page contains in the form of a waterfall. You can see whether the page violates the defined SLT, whether objects have errors, and so on. The object waterfall section displays all objects in the page in chronological order:
Object waterfall section of the Session Overview page
The object waterfall section provides the following information:
- Page start and end time—Time to load the page (equals to E2E latency)
- Object name— URL of the object as obtained by the browser
- Relative object size—Represented by circles of varying diameter. You can visually compare objects' relative sizes by using the circles, and you can recognize objects that loaded too slowly by their large size.
- Object status
- No error: blue
- With an error: red
- Type of record—Redirect (an arrow) or object (a bar)
- Time for the object to load (the length of the bar)
- Idle time—Period of inactivity between loading objects. The sum of idle times among all objects is the page idle time.
- SLT bar—Defined service level threshold for the page; visible only when the page E2E latency exceeds the defined SLT
Object Overview section
When you select an object in the object waterfall section, the information about the object appears below the waterfall:
Object Overview section of the Session Overview page
Object Overview information
|Object Latency chart|
Doughnut chart that contains the following information:
Indicates the quantity of the following types of errors and, for each error, the code and description that help you to identify the problem. To see the error description, hover over the error indicator.
Average TCP round-trip time (RTT) of all objects or pages between the client and the server, in milliseconds
Number of TCP segments retransmitted. A high number of retransmissions might indicate traffic shaping or packet loss.
Number of out-of-order TCP segments in the transmission. This is usually an indication of route flapping or poor link load balancing. This metric helps you to understand whether the objects on the page are loaded sequentially.
|Duration||Time to load the object, in milliseconds|
|Start||Time and date when the object started loading|
|Throughput||Rate of loading the object over the communication channel, in megabits per second|
|Size||Size of the object, in kilobytes|
For container objects on the page, you can drill down to back-end details by using the object trace ID. For details about the object trace ID, click the Trace Data link. When you click the link, the system redirects you to BMC Application Diagnostics.
BMC Application Diagnostics tracks only container objects. For such objects, the icon by the Trace Data link is blue. No other components of the page have a trace associated with them. A page has only one object with a trace ID.
In some cases (for example, when the object has no trace ID or BMC Application Diagnostics is not set up on the Console), the Trace Data link might be disabled. Hover over the link to see the tooltip with details about why it is disabled. For more information, contact your local Customer Support representative.
|Client IP||IP address of the user who made the request|
|Server IP||IP address of the server that responded to the user request|
Below the Object summary section, you can see the following information about object fields: