Preparation for the removal of THROW1 in SG/svn.

[loba.git] / NOTES

diff --git a/NOTES b/NOTES
index 489006c1308775df2c816b977c1969efd82dc05a..12f082bad0e45178f1369c3af61a929b84c6c345 100644 (file)
--- a/NOTES
+++ b/NOTES
@@ -1,28 +1,73 @@
-Process parameters:
+Process parameters
+==================

 
 

-    initial_load [neighbors...]
+The process parameters are: initial_load [neighbors...]

 
 

-Communications:
-     - two channels per host: normal and low_latency
+Communications
+==============

 
 

-How shall we manage link failures ?
+There are two receiving channels per host: control for information
+messages, and data for load transfers.

 
 

-Process model (?)
+Process model
+=============

 
 

-     while (there is something to do) {
-        compute some task;
-        get received tasks;
-        compute load balancing;
-        send tasks to neighbors;
-     }
+Each process is made of 3 threads: a receiver thread, a computing
+thread, and a load-balancer thread.

 
 

-    * Open Questions :
-     - definition of load on heterogeneous hosts?
-     - how to detect convergence?
-       : No need to detect convergence.  Computation stops when there
-         nothing more to do.
+* Receiver thread
+  ---------------

 
 

-     - how to manage link failures?
+    Loop
+    | wait for a message to come, either on data channel, or on ctrl channel
+    | push received message in a buffer of received messages
+    | -> ctrl messages on the one side
+    | -> data messages on the other side
+    +-

 
 

-     - shall we retrieve pending tasks?
-       : Ideally, why shall.  How?  By using some acknowledgment?
+   The loop terminates when a "finalize" message is received on each
+   channel.
+
+* Computing thread
+  ----------------
+
+    Loop
+    | if we received some real load, get it (data messages)
+    | if there is some real load to send, send it
+    | if we own some load, simulate some computing on it
+    | sleep a bit if we are looping too fast
+    +-
+    send CLOSE on data for all neighbors
+    wait for CLOSE on data from all neighbors
+
+  The loop terminates when process::still_running() returns false.
+  (read the source for full details...)
+
+* Load-balancing thread
+  ---------------------
+
+    Loop
+    | call load-balancing algorithm
+    | send ctrl messages
+    | sleep (min_lb_iter_duration)
+    | receive ctrl messages
+    +-
+    send CLOSE on ctrl for all neighbors
+    wait for CLOSE on ctrl from all neighbors
+
+  The loop terminates when process::still_running() returns false.
+  (read the source for full details...)
+
+Some notes about *load attributes
+=================================
+process::real_load              Current *real* load.
+                                Used for the computations.
+                                Displayed in log messages.
+
+process::expected_load          Current load estimation.
+                                Used for load-balancing estimation, and for
+                                diffusing to neighbors.
+                                * Without bookkeeping, it equals real_load
+                                  minus pending sends.
+                                * With bookkeeping, it corresponds to the "virtual
+                                  load".