Siege benchmarks for Pierre

Pierre Sahores sc at sahores-conseil.com
Sun Mar 27 19:30:53 EDT 2016


Richard,

If your lcHTTPd standalone is a TCP sockets server, the low TOP load (less than 25%) is not surprising at all and can even go lots less in using nginx instead of apache (6 to 8%) while the power consomption increase on the PostgreSQL controllers pool side (-> LC app’s server will never acts as a bottleneck in the global service stack). 

Best,

Pierre

Le 27 mars 2016 à 21:10, Richard Gaskin <ambassador at fourthworld.com> a écrit :
> 
> 
> Pierre, thanks for introducing me to Siege in your benchmark posts from Feb.
> 
> For those who missed them he has some great stuff there:
> http://lists.runrev.com/pipermail/use-livecode/2016-February/223352.html
> http://lists.runrev.com/pipermail/use-livecode/2016-February/223496.html
> http://lists.runrev.com/pipermail/use-livecode/2016-February/223514.html
> 
> Below are this morning's Siege runs here on a small NUC.  Each test is just delivering the default Apache page, but done in three ways: first Apache alone, then LC Server running under Apache as a CGI, then with just a LiveCode standalone handling HTTP directly:
> 
> 
> Apache 2.4.12
> -------------
> > siege http://<domain>/index.html -b -c200 -r10 -q
> Transactions:		        2000 hits
> Availability:		      100.00 %
> Elapsed time:		       29.38 secs
> Data transferred:	        6.07 MB
> Response time:		        2.49 secs
> Transaction rate:	       68.07 trans/sec
> Throughput:		        0.21 MB/sec
> Concurrency:		      169.45
> Successful transactions:        2000
> Failed transactions:	           0
> Longest transaction:	       20.78
> Shortest transaction:	        0.22
> 
> Apache + CGI, LC Server 8.0dp16
> -------------------------------
> > siege http://<domain>/tindex.lc -b -c200 -r10 -q
> Transactions:		        2000 hits
> Availability:		      100.00 %
> Elapsed time:		       53.79 secs
> Data transferred:	        6.07 MB
> Response time:		        5.22 secs
> Transaction rate:	       37.18 trans/sec
> Throughput:		        0.11 MB/sec
> Concurrency:		      194.18
> Successful transactions:        2000
> Failed transactions:	           0
> Longest transaction:	       10.42
> Shortest transaction:	        0.23
> 
> lcHTTPd standalone, made with LC 8.0dp16
> ----------------------------------------
> > siege http://<domain>:8080/index.html -b -c200 -r10 -q
> Transactions:		        2000 hits
> Availability:		      100.00 %
> Elapsed time:		       12.55 secs
> Data transferred:	        6.07 MB
> Response time:		        1.07 secs
> Transaction rate:	      159.36 trans/sec
> Throughput:		        0.48 MB/sec
> Concurrency:		      170.67
> Successful transactions:        2000
> Failed transactions:	           0
> Longest transaction:	        3.29
> Shortest transaction:	        0.07
> 
> 
> Not shown in Siege but observable in Htop is the impact on CPU: both Apache by itself and the lcHTTPd standalone run the test without ever going much above 25% load, but Apache + LC Server spikes all four cores of the N3150 in my NUC for almost the full duration of the test.
> 
> -- 
> Richard Gaskin
> Fourth World Systems
> Software Design and Development for Desktop, Mobile, and Web
> ____________________________________________________________
> Ambassador at FourthWorld.com        http://www.FourthWorld.com
> 
> _______________________________________________
> use-livecode mailing list
> use-livecode at lists.runrev.com
> Please visit this url to subscribe, unsubscribe and manage your subscription preferences:
> http://lists.runrev.com/mailman/listinfo/use-livecode

--
Pierre Sahores
mobile : 06 03 95 77 70
www.sahores-conseil.com





More information about the use-livecode mailing list