Modified:
/Introduction.wiki
=======================================
--- /Introduction.wiki Tue Jun 9 16:30:58 2009
+++ /Introduction.wiki Wed Jul 13 12:16:16 2011
@@ -1,5 +1,5 @@
+#summary Introduction to the Stajistics system
#labels Featured
-#Introduction to the Stajistics API
= What is Stajistics? =
@@ -7,7 +7,7 @@
The primary goal of Stajistics is to allow the collection of statistical
data in a program, to expose that data so that it may be queried in real
time, and to allow persistence of the data so that historical records can
be accumulated and compared. Stajistics can be used to collect data about a
running application (such as how many times a method was called), or it can
be used to collect domain-specific data (such as how many users in role "A"
are logged in).
-Stajistics is designed to be continuously active in production
environments. It tries to minimize the performance impact of statistics
collection in part by using concurrent, scalable data structures. By
default, only calculations performed on streams of data are stored in
memory, such as totals and averages, in order to keep the memory footprint
small.
+Stajistics is designed to be continuously active in production
environments. It tries to minimize the performance impact of statistics
collection in part by using concurrent, scalable data structures. By
default, only _calculations_ performed on streams of data are stored in
memory (as opposed to streams of data themselves), such as totals and
averages, in order to keep the memory footprint small.
= What is Stajistics not? =
@@ -15,7 +15,12 @@
= How can I use Stajistics? =
-TODO
+The Stajistics API is initially integrated into an application by
developers. After deployment of the application, statistics are collected
by the Stajistics system and exposed through JMX among other means.
Technical managers or system administrators may then query statistical
data, configure statistics collection, or persist data for aggregate or
trend report generation.
+
+Some example usages of Stajistics might be:
+ * Tracking hits and response times for a service
+ * Monitoring user login counts, frequencies, and durations
+ * Recording the performance of a critical algorithm
= Getting help =