X-Git-Url: https://git.xandkar.net/?p=beam_stats.git;a=blobdiff_plain;f=README.md;h=3709e2e05750d9799d2cb9aef5a9d7d62a25401c;hp=440228858d660810b0b9d96f6c9ab0d2db1be481;hb=def691558c40829e801328a4904dfed34defe186;hpb=eb6b9342c34b63de7e57493e0cc682e97046fcca

diff --git a/README.md b/README.md
index 4402288..3709e2e 100644
--- a/README.md
+++ b/README.md
@@ -21,6 +21,20 @@ Essentially like `folsomite`, but different. Different in the following ways:
     + Well-defined metrics-to-binary conversions, as opposed to the
       nearly-arbitrary term-to-string conversions used in `folsomite`
     + Spec'd, tested and Dialyzed
+- More detailed stats:
+    - **per-process**. As much process ancestry is collected as possible, then
+      anonymous processes are aggregated to their youngest-known, named
+      predecessor - this aggregation keeps the useful breadcrumbs, while
+      reducing the number of unique names from exploding, which
+      **avoids the associated problems**:
+        1. not very useful when there're lots of short-lived processes
+        2. exploading disk space usage in Whisper
+    - per-ETS-table
+
+For an example of using pre-process stats to track-down memory leaks, here's a
+screenshot of the SSL connection process memory usage growth drop after upgrade
+from 17.5 to 18.1 (back in 2015):
+![SSL memory leak going away](screenshot--2015-10-05--18.41.30.jpg)
 
 ### Adding consumers