Changes between Version 4 and Version 5 of PJNATH_Memory_Usage

Timestamp:

Jun 5, 2010 8:43:34 AM (14 years ago)

Author:

bennylp

Comment:

Added RTCP disabling optimization

PJNATH_Memory_Usage

@@ -5 +5 @@
 }}}
 
-= PJNATH Heap Usage Analysis and Optimization =
+= PJNATH/ICE Heap Usage Analysis and Optimization =
 
 '''Table of Contents'''
@@ -205 +205 @@
 [[BR]]
 
-== Optimized Results ==
+== More Optimized Results ==
 
 The result, after using the {{{config_site.h}}} settings above:
@@ -219 +219 @@
 [[BR]]
 
+
+[[BR]]
+
+== Wait, There's More! ==
+
+If memory constraint is really really tight, there is one more final optimization that we can do, i.e. '''disabling RTCP''', by declaring this macro in {{{config_site.h}}}:
+
+ {{{
+#define PJMEDIA_ADVERTISE_RTCP                  0
+ }}}
+
+Since many ICE objects are duplicated across ICE components (RTCP is an ICE component), this could potentially lower the heap usage by half!
+
+While the library currently only provides RTCP for media statistics to assist troubleshooting, still it's quite useful sometimes. You will loose RTT and TX statistics if you disable RTCP (for TX stats, you could get it in the remote endpoint of course). The system designer would need to decide whether this is a feasible optimization.
+
+== Final Result ==
+
+With RTCP '''turned off''', here are the final result:
+
+|| ||   Used || Allocated ||    Utilization% ||
+||1) App initialization, after TURN allocation  ||11,264        ||13,184        ||85||
+||2) After pjsua_start()        ||14,216        ||16,768        ||85||
+||3) Idle after initialization  ||8,304 ||9,600 ||87||
+||4) Right after making outgoing call   ||12,800        ||14,464        ||88||
+||5) After ICE negotiation is complete  ||18,544        ||20,992        ||88||
+||6) 1 minute into call ||13,136        ||14,720        ||89||
+
+It does reduce the heap consumption by close to half in some checkpoints (e.g. when idling after initialization), and significantly reduce the usages on other checkpoints.
+
+
 == Conclusion ==
 
-After the optimization, peak memory usage is around 29KB per call, compared to 75KB previously. I think that's not bad! But please continue reading the warnings below. ;-)
+We've shown that with the default settings, the '''peak''' heap usage per call is around '''76 KB''', then we reduce it to around '''29 KB''', then after the final tweak, it's down to around '''21 KB''' only. I think that's not bad! 
+
+But please continue reading the warnings below.
 
 
@@ -231 +263 @@
  - the number of candidates will vary on each host, hence the memory usages will vary.
  - these are just crude experimentations, just to give an idea on how to experiment further
+ - once again, please bear in mind that we're only optimizing PJNATH here, other settings are left to their default values.