Problem Resolution and Avoidance Analytics
The Problem Resolution and Avoidance analytics are designed to help you identify and resolve issues that occur in your infrastructure.
The Event Advisor analytic provides a prioritized list of potentially interesting events by finding spikes, relevant performance issues, and anomalous behavior, that require attention. The resulting list is ranked by magnitude and duration. Issue events can be transferred to Trend Matcher for further analysis. Event Advisor can be used as a starting point to find the root cause of an issue. The user can cast a wide net, looking for spikes or events that may shed light on where to start looking.
Trend Matcher assists you in root cause analysis by providing a way to identify the impact that entities have on each other in your infrastructure. For example, application latency in your data center can be due to any number of seemingly unrelated events. These can occur in silos (or domains) like servers, HBAs and NICs in servers, the SAN fabric, ports on your SAN attached or NAS storage array, etc. Using Trend Matcher, you can quickly and easily analyze your infrastructure to identify events or issues that correlate with a problem in your infrastructure.
Trend Matcher can be used in conjunction with Event Advisor or standalone.
Event Advisor
The Event Advisor analytic provides a prioritized list of potentially interesting events by finding spikes, relevant performance issues, and anomalous behavior, that require attention. The resulting list is ranked by magnitude and duration. Issue events can be transferred to Trend Matcher for further analysis. Event Advisor can be used as a starting point to find the root cause of an issue. The user can cast a wide net, looking for spikes or events that may shed light on where to start looking.
In this example, we'll look for events on storage ports where buffer to buffer credits are high then use Trend Matcher to find correlating entities and events to help you troubleshoot the issue.
Running Event Advisor
Start by selecting the entity type (Storage Port) and metric (% Time at Zero Transmit Credits).
Select a date range to search for buffer credit events.
Understanding Event Advisor Results
Event Advisor shows you a list of entities with interesting events, sorted by their severity, with 1 being the highest severity. Drill down on an entity to view the top ten events associated with the entity and metric.
Select Trend Match to analyze the event in more detail using the Trend Matcher analytic.
Trend Matcher
Trend Matcher accepts a source trend identified by the user or by Event Advisor and a target metric. Trend Matcher uses intelligent trend matching to compare a base trend with other entities and metrics, for the timeframe specified, and provides a topology view showing the connected entities and trends that correlate with the base trend.
This helps you identify the source of a recognized issue. For example, if there is time spent at zero Buffer-to-Buffer Credits, the Trend Matcher could find an HBA that gets busy when the buffer credit problem starts and finishes its work when the buffer credit issue goes away.
Running Trend Matcher
Trend Matcher accepts a base trend identified by the user or by the Event Advisor and a target metric. Trend Matcher uses intelligent trend matching to compare a base trend with other entities and metrics, for the time frame specified, and provides a topology view showing the connected entities and trends that correlate with the base trend.
The analysis mode offers a choice between Quick Mode and Robust Mode. Quick Mode lets you perform a quick search to find large correlations while Robust Mode provides a more comprehensive search but is slower.
There are multiple options for conducting the trend matching exercise. These options allow you to search for matches using a suggested or targeted situation, to limit the search to only connected entities, or search for matches across all entities.
Understanding Trend Matcher Results
The Trend Matcher results are divided into two sections: A Topology tab showing you where entities and metrics correlated with the base trend, and a list of correlations found that match the base trend.
The solid blue circles indicate where there were matches. The circles around the solid circle indicate the degree of correlation: the more blue circles displayed around an entity, the higher the correlation. Entities displaying three dots at their center indicate a correlation at the sub-entity level – drill down to view more details.
To the right of the topology map is a list of all matching entities and events, ordered by percent correlation. The list displays both positive and negative correlations.
Expand the correlated entity and metric to see how it compares with the base trend. The correlated entity is highlighted on the topology map to the left.
This gives you valuable information for analyzing the issue. For example, a flow control issue may be related to increased workload on an HBA port. Once the workload decreases, the buffer credit issue on the storage port also resolves.
You can also run Trend Matcher as a standalone analytic. You'll need to select a base trend to use for the analysis.