TM module tests

Standard tests done against public SER version. SER used 512 MB of shared memory.

throughput

sl_reply
Statelesly respond received message to see the limit.

forward
Forward received message stateless to see differences between stateless and statefull forwarding.

t_reply.1
This scenario uses TM to reply received message with t_reply.

This is 1. iteration.

t_relay.1
Forward received message with t_relay.

This is 1. iteration.

t_reply.2
This scenario uses TM to reply received message with t_reply.

This is 2. iteration.

t_relay.2
Forward received message with t_relay.

This is 2. iteration.


sl_reply

Statelesly respond received message to see the limit.

additional SIPP args:

Used config file was sl_reply2.cfg (log here), sipp scenario message.xml, sipp output screens are in sl_reply.screen.

SAR statistics: platform dependent binary data, average values, all values.

SIPP command

sipp -nd -sf scenarios/message.xml -trace_screen -trace_rtt -domain "213.192.30.237" -r 20000 -l 32 -m 5000000 213.192.30.219:5060

Test stopped after 5000000 iterations, maximum allowed simultaneous calls were 32; reached average call throughput was 11115 calls per second.

Iterations 5000000
Max concurent calls 32
Max allowed Rate 20000 requests per second
Average Call Rate 11115 requests per second
Failures 0
Retransmissions 0
Timeout 0
Elapsed Time 00:07:29:844 hour:min:sec:msec

response times response times - histogram CPU usage request times - histogram Shared memory usage


forward

Forward received message stateless to see differences between stateless and statefull forwarding.

additional SIPP args: -test forward

Used config file was tm.cfg (log here), sipp scenario register_users.xml, sipp output screens are in forward.screen.

SAR statistics: platform dependent binary data, average values, all values.

Note

SER restart disabled - using running SER instance...

SIPP command

sipp -nd -sf scenarios/register_users.xml -trace_screen -trace_rtt -domain "213.192.30.237" -r 20000 -l 32 -m 5000000 213.192.30.219:5060 -test forward

Test stopped after 5000000 iterations, maximum allowed simultaneous calls were 32; reached average call throughput was 10802 calls per second.

Iterations 5000000
Max concurent calls 32
Max allowed Rate 20000 requests per second
Average Call Rate 10802 requests per second
Failures 0
Retransmissions 346
Timeout 0
Elapsed Time 00:07:42:866 hour:min:sec:msec

response times response times - histogram CPU usage request times - histogram Shared memory usage


t_reply.1

This scenario uses TM to reply received message with t_reply.

This is 1. iteration.

additional SIPP args: -test reply

Used config file was tm.cfg (log here), sipp scenario register_users.xml, sipp output screens are in t_reply.1.screen.

SAR statistics: platform dependent binary data, average values, all values.

Note

SER restart disabled - using running SER instance...

SIPP command

sipp -nd -sf scenarios/register_users.xml -trace_screen -trace_rtt -domain "example.com" -r 20000 -l 32 -m 5000000 213.192.30.219:5060 -test reply

Test stopped after 5000000 iterations, maximum allowed simultaneous calls were 32; reached average call throughput was 9572 calls per second.

Iterations 5000000
Max concurent calls 32
Max allowed Rate 20000 requests per second
Average Call Rate 9572 requests per second
Failures 0
Retransmissions 559
Timeout 0
Elapsed Time 00:08:42:371 hour:min:sec:msec

response times response times - histogram CPU usage request times - histogram Shared memory usage


t_relay.1

Forward received message with t_relay.

This is 1. iteration.

additional SIPP args: -test relay

Used config file was tm.cfg (log here), sipp scenario register_users.xml, sipp output screens are in t_relay.1.screen.

SAR statistics: platform dependent binary data, average values, all values.

Note

SER restart disabled - using running SER instance...

SIPP command

sipp -nd -sf scenarios/register_users.xml -trace_screen -trace_rtt -domain "213.192.30.237" -r 20000 -l 32 -m 5000000 213.192.30.219:5060 -test relay

Test stopped after 5000000 iterations, maximum allowed simultaneous calls were 32; reached average call throughput was 10664 calls per second.

Iterations 5000000
Max concurent calls 32
Max allowed Rate 20000 requests per second
Average Call Rate 10664 requests per second
Failures 0
Retransmissions 54
Timeout 0
Elapsed Time 00:07:48:888 hour:min:sec:msec

response times response times - histogram CPU usage request times - histogram Shared memory usage


t_reply.2

This scenario uses TM to reply received message with t_reply.

This is 2. iteration.

additional SIPP args: -test reply

Used config file was tm.cfg (log here), sipp scenario register_users.xml, sipp output screens are in t_reply.2.screen.

SAR statistics: platform dependent binary data, average values, all values.

Note

SER restart disabled - using running SER instance...

SIPP command

sipp -nd -sf scenarios/register_users.xml -trace_screen -trace_rtt -domain "example.com" -r 20000 -l 32 -m 5000000 213.192.30.219:5060 -test reply

Test stopped after 5000000 iterations, maximum allowed simultaneous calls were 32; reached average call throughput was 9402 calls per second.

Iterations 5000000
Max concurent calls 32
Max allowed Rate 20000 requests per second
Average Call Rate 9402 requests per second
Failures 0
Retransmissions 23
Timeout 0
Elapsed Time 00:08:51:786 hour:min:sec:msec

response times response times - histogram CPU usage request times - histogram Shared memory usage


t_relay.2

Forward received message with t_relay.

This is 2. iteration.

additional SIPP args: -test relay

Used config file was tm.cfg (log here), sipp scenario register_users.xml, sipp output screens are in t_relay.2.screen.

SAR statistics: platform dependent binary data, average values, all values.

Note

SER restart disabled - using running SER instance...

SIPP command

sipp -nd -sf scenarios/register_users.xml -trace_screen -trace_rtt -domain "213.192.30.237" -r 20000 -l 32 -m 5000000 213.192.30.219:5060 -test relay

Test stopped after 5000000 iterations, maximum allowed simultaneous calls were 32; reached average call throughput was 10627 calls per second.

Iterations 5000000
Max concurent calls 32
Max allowed Rate 20000 requests per second
Average Call Rate 10627 requests per second
Failures 0
Retransmissions 0
Timeout 0
Elapsed Time 00:07:50:504 hour:min:sec:msec

response times response times - histogram CPU usage request times - histogram Shared memory usage


Test information

Test done by kubartv from jin on Wed Mar 28 14:38:25 CEST 2007.

SIPP version:
Sipp v1.1, version 20060829, built Mar 28 2007, 11:22:45. This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA Author: see source files.

Tested machine

Test done under user kubartv on 213.192.30.219, helping machine (if used) is 213.192.30.237, user kubartv.

Linux pres 2.6.17-2-686-bigmem #1 SMP Sat Sep 23 21:15:25 UTC 2006 i686 GNU/Linux

CPU

processor	: 0
vendor_id	: GenuineIntel
cpu family	: 15
model		: 4
model name	: Intel(R) Xeon(TM) CPU 3.00GHz
stepping	: 3
cpu MHz		: 3000.621
cache size	: 2048 KB
physical id	: 0
siblings	: 2
core id		: 0
cpu cores	: 1
fdiv_bug	: no
hlt_bug		: no
f00f_bug	: no
coma_bug	: no
fpu		: yes
fpu_exception	: yes
cpuid level	: 5
wp		: yes
flags		: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm constant_tsc pni monitor ds_cpl cid cx16 xtpr
bogomips	: 6006.61

processor	: 1
vendor_id	: GenuineIntel
cpu family	: 15
model		: 4
model name	: Intel(R) Xeon(TM) CPU 3.00GHz
stepping	: 3
cpu MHz		: 3000.621
cache size	: 2048 KB
physical id	: 0
siblings	: 2
core id		: 0
cpu cores	: 1
fdiv_bug	: no
hlt_bug		: no
f00f_bug	: no
coma_bug	: no
fpu		: yes
fpu_exception	: yes
cpuid level	: 5
wp		: yes
flags		: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm constant_tsc pni monitor ds_cpl cid cx16 xtpr
bogomips	: 6000.68

Memory

             total       used       free     shared    buffers     cached
Mem:       8312532    3490184    4822348          0     163956    3134664
-/+ buffers/cache:     191564    8120968
Swap:      7807352          0    7807352

SER version

version: ser 2.0.0-rc1 (i386/linux) flags: STATS: Off, USE_IPV6, USE_TCP, USE_TLS, TLS_HOOKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MEM, SHM_MMAP, PKG_MALLOC, DBG_QM_MALLOC, FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_DST_BLACKLIST ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16, MAX_URI_SIZE 1024, BUF_SIZE 65535 poll method support: poll, epoll_lt, epoll_et, sigio_rt, select. @(#) $Id: main.c,v 1.228.2.2 2007/03/01 13:53:37 andrei Exp $ main.c compiled on 10:34:32 Mar 26 2007 with gcc 3.3.5

Graph descriptions

There are these graphs for each scenario:

response delay
Time for each response; on X-axis is request time, i.e. when was the request sent, on Y-axis is how long time was between sending request and receiving the response. This value is not the same as values measured by watching network traffic probably due to sipp processing delay.

response delay histogram
Histogram made from time differences between sending request and receiving response. It shows numbers of response times grouped by 10 ms.

CPU usage
Shows sumary for all monitored processes for given time. Monitored processors can be set within variables script when generating the output (not needed in runtime). This output is taken from top which collects this information for one second. It can be more than 100% in the case of more processors (for example tc4 has two hyperthreaded processors thus it can show values up to 400%).

sent request histogram
Shows numbers of sent requests grouped by one second. Sometimes it shows strange jumps (for example in t_relay scenario after shmem is gone)...

Test result notes

SIPP is set to allow at most 32 calls at the same time. This means that when there are 32 pending calls (waiting for response) it doesn't create new ones.