Factorize computation of load_to_send.

[loba.git] / NOTES

diff --git a/NOTES b/NOTES
index 5399a4c982987d69a584edd45bda82d6ba92be21..71e27296cc07fbd0dd958b5b016b92b4f8ea620e 100644 (file)
--- a/NOTES
+++ b/NOTES
@@ -12,11 +12,11 @@ messages, and data for load transfers.
 Process model
 =============
 
-Each process is made of 3 threads: a receiver thread, a computing
-thread, and a load-balancer thread.
+Each process is made of 3 threads: a receiving thread, a computing
+thread, and a load-balancing thread.
 
-* Receiver thread
-  ---------------
+* Receiving thread
+  ----------------
 
     Loop
     | wait for a message to come, either on data channel, or on ctrl channel
@@ -47,10 +47,10 @@ thread, and a load-balancer thread.
   ---------------------
 
     Loop
+    | receive ctrl messages
     | 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
@@ -71,3 +71,7 @@ process::expected_load          Current load estimation.
                                   minus pending sends.
                                 * With bookkeeping, it corresponds to the
                                   "virtual load".
+
+process::received_load          Real load received from neighbors.
+                                Used when receiveing data messages, and then
+                                added to real_load.