SIP performance tweaking
Hi, I am currently doing some characterization testing to figure out the limitations of my freeswitch server. I have tried both with SIPP and by driving SIP calls between 2 Freeswitch through mod_event_socket. I get similar results both ways and I am suspicious of the results I get. Basically, whenever I try to launch calls with SIPP at a rate superior than 8 cps, it seems like calls are being stacked. The response time for a call between the "invite" and the corresponding "200 OK" keeps increasing as long as calls are being made which slows down the processing speed until the average CPS gets around 8 cps. My setup is : 1 Dual-Core AMD Opteron(tm) Processor 8216 @ 2.4 GHz 4 GB of RAM Gigabit NIC Centos 5 FreeSwitch Version 1.0.trunk (8911M) I think that my setup is underperforming and that it might be related to some missing configuration tweaks. I already adjusted the ulimit to the recommended settings in the wiki for maximum performance. I was wondering if there might be something else, that I'm still missing. Would any of you have an idea of what might be holding back my setup ? Thanks ! PG P.S. I gathered a bunch of information that I thought could be useful. Freeswitch performance settings In switch.conf.xml <!--Most channels to allow at once --> <param name="max-sessions" value="10000"/> <!--Most channels to create per second --> <param name="sessions-per-second" value="1000"/> <!-- Default Global Log Level - value is one of debug,info,notice,warning,err,crit,alert --> <param name="loglevel" value="alert"/> Here is the SIPP command I used, based on Freeswitch load_test script. # sipp -sf dft_cap.xml 192.168.100.9 -s 9990 -r 15 -rp 1000 -l 200 -d 10000 -i 192.168.100.200 -mi 192.168.100.200 # cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 65 model name : Dual-Core AMD Opteron(tm) Processor 8216 stepping : 2 cpu MHz : 2411.121 cache size : 1024 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy bogomips : 4823.60 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc # cat /proc/meminfo MemTotal: 4042672 kB MemFree: 843188 kB Buffers: 256676 kB Cached: 2671172 kB SwapCached: 1092 kB Active: 777140 kB Inactive: 2192840 kB HighTotal: 0 kB HighFree: 0 kB LowTotal: 4042672 kB LowFree: 843188 kB SwapTotal: 2031608 kB SwapFree: 2016120 kB Dirty: 0 kB Writeback: 0 kB AnonPages: 41068 kB Mapped: 26028 kB Slab: 195536 kB PageTables: 5672 kB NFS_Unstable: 0 kB Bounce: 0 kB CommitLimit: 4052944 kB Committed_AS: 153064 kB VmallocTotal: 34359738367 kB VmallocUsed: 35492 kB VmallocChunk: 34359692283 kB HugePages_Total: 0 HugePages_Free: 0 HugePages_Rsvd: 0 Hugepagesize: 2048 kB # ulimit -a core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited scheduling priority (-e) 0 file size (blocks, -f) unlimited pending signals (-i) 40960 max locked memory (kbytes, -l) 32 max memory size (kbytes, -m) unlimited open files (-n) 1024 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) 819200 real-time priority (-r) 0 stack size (kbytes, -s) 10240 cpu time (seconds, -t) unlimited max user processes (-u) 40960 virtual memory (kbytes, -v) unlimited file locks (-x) unlimited # top top - 14:31:16 up 1 day, 58 min, 2 users, load average: 5.73, 6.83, 5.36 Tasks: 129 total, 1 running, 128 sleeping, 0 stopped, 0 zombie Cpu(s): 7.0%us, 2.8%sy, 0.0%ni, 50.8%id, 38.8%wa, 0.0%hi, 0.5%si, 0.0%st Mem: 4042672k total, 3418968k used, 623704k free, 256904k buffers Swap: 2031608k total, 15488k used, 2016120k free, 2681820k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 6078 root 18 0 614m 109m 6480 S 19 2.8 3:39.11 freeswitch
what is the callflow of the extension you are calling with sipp? start by making the target extension just answer and play back a file till you have it working and expand it from there. you can comment out the manage-presence option in your sip profile to make the sip a lot less resource intensive. here are the ulimits I set for load testing ulimit -c unlimited ulimit -d unlimited ulimit -f unlimited ulimit -i unlimited ulimit -n 999999 ulimit -q unlimited ulimit -u unlimited ulimit -v unlimited ulimit -x unlimited ulimit -s 244 ulimit -l unlimited On Tue, Jul 8, 2008 at 1:51 PM, Patrick Grondin <pgrondin@ip5.com> wrote:
Hi,
I am currently doing some characterization testing to figure out the limitations of my freeswitch server. I have tried both with SIPP and by driving SIP calls
between 2 Freeswitch through mod_event_socket. I get similar results both ways and I am suspicious of the results I get.
Basically, whenever I try to launch calls with SIPP at a rate superior than 8 cps, it seems like calls are being stacked. The response time for a call between
the "invite" and the corresponding "200 OK" keeps increasing as long as calls are being made which slows down the processing speed until the average
CPS gets around 8 cps.
My setup is :
1 Dual-Core AMD Opteron(tm) Processor 8216 @ 2.4 GHz
4 GB of RAM
Gigabit NIC
Centos 5
FreeSwitch Version 1.0.trunk (8911M)
I think that my setup is underperforming and that it might be related to some missing configuration tweaks. I already adjusted the ulimit to the
recommended settings in the wiki for maximum performance. I was wondering if there might be something else, that I'm still missing.
Would any of you have an idea of what might be holding back my setup ?
Thanks !
PG
P.S. I gathered a bunch of information that I thought could be useful.
Freeswitch performance settings In switch.conf.xml
<!--Most channels to allow at once -->
<param name="max-sessions" value="10000"/>
<!--Most channels to create per second -->
<param name="sessions-per-second" value="1000"/>
<!-- Default Global Log Level - value is one of debug,info,notice,warning,err,crit,alert -->
<param name="loglevel" value="alert"/>
Here is the SIPP command I used, based on Freeswitch load_test script.
# sipp -sf dft_cap.xml 192.168.100.9 -s 9990 -r 15 -rp 1000 -l 200 -d 10000 -i 192.168.100.200 -mi 192.168.100.200
# cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 65
model name : Dual-Core AMD Opteron(tm) Processor 8216
stepping : 2
cpu MHz : 2411.121
cache size : 1024 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy
bogomips : 4823.60
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc
# cat /proc/meminfo
MemTotal: 4042672 kB
MemFree: 843188 kB
Buffers: 256676 kB
Cached: 2671172 kB
SwapCached: 1092 kB
Active: 777140 kB
Inactive: 2192840 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 4042672 kB
LowFree: 843188 kB
SwapTotal: 2031608 kB
SwapFree: 2016120 kB
Dirty: 0 kB
Writeback: 0 kB
AnonPages: 41068 kB
Mapped: 26028 kB
Slab: 195536 kB
PageTables: 5672 kB
NFS_Unstable: 0 kB
Bounce: 0 kB
CommitLimit: 4052944 kB
Committed_AS: 153064 kB
VmallocTotal: 34359738367 kB
VmallocUsed: 35492 kB
VmallocChunk: 34359692283 kB
HugePages_Total: 0
HugePages_Free: 0
HugePages_Rsvd: 0
Hugepagesize: 2048 kB
# ulimit –a
core file size (blocks, -c) 0
data seg size (kbytes, -d) unlimited
scheduling priority (-e) 0
file size (blocks, -f) unlimited
pending signals (-i) 40960
max locked memory (kbytes, -l) 32
max memory size (kbytes, -m) unlimited
open files (-n) 1024
pipe size (512 bytes, -p) 8
POSIX message queues (bytes, -q) 819200
real-time priority (-r) 0
stack size (kbytes, -s) 10240
cpu time (seconds, -t) unlimited
max user processes (-u) 40960
virtual memory (kbytes, -v) unlimited
file locks (-x) unlimited
# top
top - 14:31:16 up 1 day, 58 min, 2 users, load average: 5.73, 6.83, 5.36
Tasks: 129 total, 1 running, 128 sleeping, 0 stopped, 0 zombie
Cpu(s): 7.0%us, 2.8%sy, 0.0%ni, 50.8%id, 38.8%wa, 0.0%hi, 0.5%si, 0.0%st
Mem: 4042672k total, 3418968k used, 623704k free, 256904k buffers
Swap: 2031608k total, 15488k used, 2016120k free, 2681820k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
6078 root 18 0 614m 109m 6480 S 19 2.8 3:39.11 freeswitch
_______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888 googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:213-799-1400
Hi, I've tried the tweaks (the ulimit changes and the disabling of sip presence) you suggested and it did improve a little, going from 8 CPS to 9.7 CPS. My callflow is real simple. I'm calling an extension that plays a wav file and I hang up the call, 10 seconds after the start of the call, before the end of the playback. You can see from the wireshark callflow below, that at a call rate of 15 CPS, after 20 seconds, it takes 17 seconds for a call to go from 100 Trying to 200 OK. Do you have any other ideas of what could be holding down my freeswitch ? 192.168.100.200 = SIPP, call initiation 192.168.100.9 = Freeswitch, call receiver |Time | 192.168.100.200 | 192.168.100.9 | |20.013 | INVITE SDP ( g711U) |SIP From: sip:sipp@192.168.100.200:5060 To:sip:9990@192.168.100.9:5060 | |(5060) ------------------> (5060) | |20.013 | 100 Trying| |SIP Status | |(5060) <------------------ (5060) | |20.014 | RTP (g711U) |RTP Num packets:8 Duration:0.061s SSRC:0xF3FF | |(6000) <------------------ (24806) | |20.014 | RTP (g711U) |RTP Num packets:16 Duration:0.060s SSRC:0xD47F | |(6000) <------------------ (25394) | |20.014 | RTP (g711U) |RTP Num packets:12 Duration:0.060s SSRC:0x1463 | |(6000) <------------------ (22286) | |20.014 | RTP (g711U) |RTP Num packets:16 Duration:0.060s SSRC:0xBCC0 | |(6000) <------------------ (22934) | |20.014 | RTP (g711U) |RTP Num packets:8 Duration:0.060s SSRC:0xC8E9 | |(6000) <------------------ (18672) | |20.014 | RTP (g711U) |RTP Num packets:16 Duration:0.060s SSRC:0xF7D | |(6000) <------------------ (16944) | |20.014 | RTP (g711U) |RTP Num packets:24 Duration:0.060s SSRC:0x5927 | |(6000) <------------------ (31390) | |20.015 | RTP (g711U) |RTP Num packets:24 Duration:0.059s SSRC:0x8352 | |(6000) <------------------ (23078) | |20.015 | RTP (g711U) |RTP Num packets:20 Duration:0.060s SSRC:0xF637 | |(6000) <------------------ (20284) | |20.015 | RTP (g711U) |RTP Num packets:12 Duration:0.058s SSRC:0xD3D6 | |(6000) <------------------ (30944) | |20.016 | RTP (g711U) |RTP Num packets:12 Duration:0.058s SSRC:0x289 | |(6000) <------------------ (20836) | |37.578 | 200 OK SDP ( CN) |SIP Status | |(5060) <------------------ (5060) | |37.579 | ACK | |SIP Request | |(5060) ------------------> (5060) | |47.579 | BYE | |SIP Request | |(5060) ------------------> (5060) | |47.580 | 200 OK | |SIP Status | |(5060) <------------------ (5060) | PG From: freeswitch-users-bounces@lists.freeswitch.org [mailto:freeswitch-users-bounces@lists.freeswitch.org] On Behalf Of Anthony Minessale Sent: July-08-08 3:52 PM To: freeswitch-users@lists.freeswitch.org Subject: Re: [Freeswitch-users] SIP performance tweaking what is the callflow of the extension you are calling with sipp? start by making the target extension just answer and play back a file till you have it working and expand it from there. you can comment out the manage-presence option in your sip profile to make the sip a lot less resource intensive. here are the ulimits I set for load testing ulimit -c unlimited ulimit -d unlimited ulimit -f unlimited ulimit -i unlimited ulimit -n 999999 ulimit -q unlimited ulimit -u unlimited ulimit -v unlimited ulimit -x unlimited ulimit -s 244 ulimit -l unlimited On Tue, Jul 8, 2008 at 1:51 PM, Patrick Grondin <pgrondin@ip5.com<mailto:pgrondin@ip5.com>> wrote: Hi, I am currently doing some characterization testing to figure out the limitations of my freeswitch server. I have tried both with SIPP and by driving SIP calls between 2 Freeswitch through mod_event_socket. I get similar results both ways and I am suspicious of the results I get. Basically, whenever I try to launch calls with SIPP at a rate superior than 8 cps, it seems like calls are being stacked. The response time for a call between the "invite" and the corresponding "200 OK" keeps increasing as long as calls are being made which slows down the processing speed until the average CPS gets around 8 cps. My setup is : 1 Dual-Core AMD Opteron(tm) Processor 8216 @ 2.4 GHz 4 GB of RAM Gigabit NIC Centos 5 FreeSwitch Version 1.0.trunk (8911M) I think that my setup is underperforming and that it might be related to some missing configuration tweaks. I already adjusted the ulimit to the recommended settings in the wiki for maximum performance. I was wondering if there might be something else, that I'm still missing. Would any of you have an idea of what might be holding back my setup ? Thanks ! PG P.S. I gathered a bunch of information that I thought could be useful. Freeswitch performance settings In switch.conf.xml <!--Most channels to allow at once --> <param name="max-sessions" value="10000"/> <!--Most channels to create per second --> <param name="sessions-per-second" value="1000"/> <!-- Default Global Log Level - value is one of debug,info,notice,warning,err,crit,alert --> <param name="loglevel" value="alert"/> Here is the SIPP command I used, based on Freeswitch load_test script. # sipp -sf dft_cap.xml 192.168.100.9<http://192.168.100.9> -s 9990 -r 15 -rp 1000 -l 200 -d 10000 -i 192.168.100.200<http://192.168.100.200> -mi 192.168.100.200<http://192.168.100.200> # cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 65 model name : Dual-Core AMD Opteron(tm) Processor 8216 stepping : 2 cpu MHz : 2411.121 cache size : 1024 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy bogomips : 4823.60 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc # cat /proc/meminfo MemTotal: 4042672 kB MemFree: 843188 kB Buffers: 256676 kB Cached: 2671172 kB SwapCached: 1092 kB Active: 777140 kB Inactive: 2192840 kB HighTotal: 0 kB HighFree: 0 kB LowTotal: 4042672 kB LowFree: 843188 kB SwapTotal: 2031608 kB SwapFree: 2016120 kB Dirty: 0 kB Writeback: 0 kB AnonPages: 41068 kB Mapped: 26028 kB Slab: 195536 kB PageTables: 5672 kB NFS_Unstable: 0 kB Bounce: 0 kB CommitLimit: 4052944 kB Committed_AS: 153064 kB VmallocTotal: 34359738367 kB VmallocUsed: 35492 kB VmallocChunk: 34359692283 kB HugePages_Total: 0 HugePages_Free: 0 HugePages_Rsvd: 0 Hugepagesize: 2048 kB # ulimit -a core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited scheduling priority (-e) 0 file size (blocks, -f) unlimited pending signals (-i) 40960 max locked memory (kbytes, -l) 32 max memory size (kbytes, -m) unlimited open files (-n) 1024 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) 819200 real-time priority (-r) 0 stack size (kbytes, -s) 10240 cpu time (seconds, -t) unlimited max user processes (-u) 40960 virtual memory (kbytes, -v) unlimited file locks (-x) unlimited # top top - 14:31:16 up 1 day, 58 min, 2 users, load average: 5.73, 6.83, 5.36 Tasks: 129 total, 1 running, 128 sleeping, 0 stopped, 0 zombie Cpu(s): 7.0%us, 2.8%sy, 0.0%ni, 50.8%id, 38.8%wa, 0.0%hi, 0.5%si, 0.0%st Mem: 4042672k total, 3418968k used, 623704k free, 256904k buffers Swap: 2031608k total, 15488k used, 2016120k free, 2681820k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 6078 root 18 0 614m 109m 6480 S 19 2.8 3:39.11 freeswitch _______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org<mailto:Freeswitch-users@lists.freeswitch.org> http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org -- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ AIM: anthm MSN:anthony_minessale@hotmail.com<mailto:MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<mailto:PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net<http://irc.freenode.net> #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org<mailto:sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888<http://iax:guest@conference.freeswitch.org/888> googletalk:conf+888@conference.freeswitch.org<mailto:googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:213-799-1400
Can you cat /proc/cpuinfo and uname -a ? /b On Jul 9, 2008, at 2:49 PM, Patrick Grondin wrote:
Hi,
I’ve tried the tweaks (the ulimit changes and the disabling of sip presence) you suggested and it did improve a little, going from 8 CPS to 9.7 CPS.
My callflow is real simple. I’m calling an extension that plays a wav file and I hang up the call, 10 seconds after the start of the call, before the end of the playback.
You can see from the wireshark callflow below, that at a call rate of 15 CPS, after 20 seconds, it takes 17 seconds for a call to go from 100 Trying to 200 OK.
Do you have any other ideas of what could be holding down my freeswitch ?
Brian West sip:brian@freeswitch.org
# ulimit -a core file size (blocks, -c) unlimited data seg size (kbytes, -d) unlimited scheduling priority (-e) 0 file size (blocks, -f) unlimited pending signals (-i) unlimited max locked memory (kbytes, -l) unlimited max memory size (kbytes, -m) unlimited open files (-n) 999999 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) unlimited real-time priority (-r) 0 stack size (kbytes, -s) 244 cpu time (seconds, -t) unlimited max user processes (-u) unlimited virtual memory (kbytes, -v) unlimited file locks (-x) unlimited #cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 65 model name : Dual-Core AMD Opteron(tm) Processor 8216 stepping : 2 cpu MHz : 2411.121 cache size : 1024 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy bogomips : 4823.60 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc processor : 1 vendor_id : AuthenticAMD cpu family : 15 model : 65 model name : Dual-Core AMD Opteron(tm) Processor 8216 stepping : 2 cpu MHz : 2411.121 cache size : 1024 KB physical id : 0 siblings : 2 core id : 1 cpu cores : 2 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy bogomips : 4821.58 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc From: freeswitch-users-bounces@lists.freeswitch.org [mailto:freeswitch-users-bounces@lists.freeswitch.org] On Behalf Of Brian West Sent: July-09-08 3:57 PM To: freeswitch-users@lists.freeswitch.org Subject: Re: [Freeswitch-users] SIP performance tweaking Can you cat /proc/cpuinfo and uname -a ? /b On Jul 9, 2008, at 2:49 PM, Patrick Grondin wrote: Hi, I've tried the tweaks (the ulimit changes and the disabling of sip presence) you suggested and it did improve a little, going from 8 CPS to 9.7 CPS. My callflow is real simple. I'm calling an extension that plays a wav file and I hang up the call, 10 seconds after the start of the call, before the end of the playback. You can see from the wireshark callflow below, that at a call rate of 15 CPS, after 20 seconds, it takes 17 seconds for a call to go from 100 Trying to 200 OK. Do you have any other ideas of what could be holding down my freeswitch ? Brian West sip:brian@freeswitch.org
Sorry my mistake ;) # uname -a Linux tyan-centOS64 2.6.18-92.1.6.el5 #1 SMP Wed Jun 25 13:45:47 EDT 2008 x86_64 x86_64 x86_64 GNU/Linux From: freeswitch-users-bounces@lists.freeswitch.org [mailto:freeswitch-users-bounces@lists.freeswitch.org] On Behalf Of Brian West Sent: July-09-08 4:06 PM To: freeswitch-users@lists.freeswitch.org Subject: Re: [Freeswitch-users] SIP performance tweaking uname -a ;) On Jul 9, 2008, at 3:02 PM, Patrick Grondin wrote: # ulimit -a Brian West sip:brian@freeswitch.org
can you join irc and let us in the box to examine it. all of this email back and forth is not going to work. There is an applet on our homepage to join or use any irc client to get to irc.freenode.net #freeswitch if you don't have irc use one of my im addrs in my sig to contact me. On Wed, Jul 9, 2008 at 3:02 PM, Patrick Grondin <pgrondin@ip5.com> wrote:
# ulimit –a
core file size (blocks, -c) unlimited
data seg size (kbytes, -d) unlimited
scheduling priority (-e) 0
file size (blocks, -f) unlimited
pending signals (-i) unlimited
max locked memory (kbytes, -l) unlimited
max memory size (kbytes, -m) unlimited
open files (-n) 999999
pipe size (512 bytes, -p) 8
POSIX message queues (bytes, -q) unlimited
real-time priority (-r) 0
stack size (kbytes, -s) 244
cpu time (seconds, -t) unlimited
max user processes (-u) unlimited
virtual memory (kbytes, -v) unlimited
file locks (-x) unlimited
#cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 65
model name : Dual-Core AMD Opteron(tm) Processor 8216
stepping : 2
cpu MHz : 2411.121
cache size : 1024 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy
bogomips : 4823.60
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc
processor : 1
vendor_id : AuthenticAMD
cpu family : 15
model : 65
model name : Dual-Core AMD Opteron(tm) Processor 8216
stepping : 2
cpu MHz : 2411.121
cache size : 1024 KB
physical id : 0
siblings : 2
core id : 1
cpu cores : 2
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy
bogomips : 4821.58
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc
*From:* freeswitch-users-bounces@lists.freeswitch.org [mailto: freeswitch-users-bounces@lists.freeswitch.org] *On Behalf Of *Brian West *Sent:* July-09-08 3:57 PM *To:* freeswitch-users@lists.freeswitch.org *Subject:* Re: [Freeswitch-users] SIP performance tweaking
Can you cat /proc/cpuinfo and uname -a ?
/b
On Jul 9, 2008, at 2:49 PM, Patrick Grondin wrote:
Hi,
I've tried the tweaks (the ulimit changes and the disabling of sip presence) you suggested and it did improve a little, going from 8 CPS to 9.7 CPS.
My callflow is real simple. I'm calling an extension that plays a wav file and I hang up the call, 10 seconds after the start of the call, before the end of the playback.
You can see from the wireshark callflow below, that at a call rate of 15 CPS, after 20 seconds, it takes 17 seconds for a call to go from 100 Trying to 200 OK.
Do you have any other ideas of what could be holding down my freeswitch ?
Brian West
sip:brian@freeswitch.org
_______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888 googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:213-799-1400
Just a follow up on the new results I got while testing. I had to do a couple of tests to figure if my problem was related to my code or the problem anthm figured out. After a couple of tests, I can say that my problem was mostly related to the way my dialplan was made. The extensions I had added to my dialplan where at the very end of the default dialplan, so for every call I made to these extensions, it had to go through the whole default dialplan which slowed down things. After putting my test extensions at the top of the file, I can gracefully achieve 30 CPS and I can see that I'm CPU bound to get more performance. From: freeswitch-users-bounces@lists.freeswitch.org [mailto:freeswitch-users-bounces@lists.freeswitch.org] On Behalf Of Anthony Minessale Sent: July-09-08 4:15 PM To: freeswitch-users@lists.freeswitch.org Subject: Re: [Freeswitch-users] SIP performance tweaking can you join irc and let us in the box to examine it. all of this email back and forth is not going to work. There is an applet on our homepage to join or use any irc client to get to irc.freenode.net<http://irc.freenode.net> #freeswitch if you don't have irc use one of my im addrs in my sig to contact me. On Wed, Jul 9, 2008 at 3:02 PM, Patrick Grondin <pgrondin@ip5.com<mailto:pgrondin@ip5.com>> wrote: # ulimit -a core file size (blocks, -c) unlimited data seg size (kbytes, -d) unlimited scheduling priority (-e) 0 file size (blocks, -f) unlimited pending signals (-i) unlimited max locked memory (kbytes, -l) unlimited max memory size (kbytes, -m) unlimited open files (-n) 999999 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) unlimited real-time priority (-r) 0 stack size (kbytes, -s) 244 cpu time (seconds, -t) unlimited max user processes (-u) unlimited virtual memory (kbytes, -v) unlimited file locks (-x) unlimited #cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 65 model name : Dual-Core AMD Opteron(tm) Processor 8216 stepping : 2 cpu MHz : 2411.121 cache size : 1024 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy bogomips : 4823.60 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc processor : 1 vendor_id : AuthenticAMD cpu family : 15 model : 65 model name : Dual-Core AMD Opteron(tm) Processor 8216 stepping : 2 cpu MHz : 2411.121 cache size : 1024 KB physical id : 0 siblings : 2 core id : 1 cpu cores : 2 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy bogomips : 4821.58 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc From: freeswitch-users-bounces@lists.freeswitch.org<mailto:freeswitch-users-bounces@lists.freeswitch.org> [mailto:freeswitch-users-bounces@lists.freeswitch.org<mailto:freeswitch-users-bounces@lists.freeswitch.org>] On Behalf Of Brian West Sent: July-09-08 3:57 PM To: freeswitch-users@lists.freeswitch.org<mailto:freeswitch-users@lists.freeswitch.org> Subject: Re: [Freeswitch-users] SIP performance tweaking Can you cat /proc/cpuinfo and uname -a ? /b On Jul 9, 2008, at 2:49 PM, Patrick Grondin wrote: Hi, I've tried the tweaks (the ulimit changes and the disabling of sip presence) you suggested and it did improve a little, going from 8 CPS to 9.7 CPS. My callflow is real simple. I'm calling an extension that plays a wav file and I hang up the call, 10 seconds after the start of the call, before the end of the playback. You can see from the wireshark callflow below, that at a call rate of 15 CPS, after 20 seconds, it takes 17 seconds for a call to go from 100 Trying to 200 OK. Do you have any other ideas of what could be holding down my freeswitch ? Brian West sip:brian@freeswitch.org _______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org<mailto:Freeswitch-users@lists.freeswitch.org> http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org -- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ AIM: anthm MSN:anthony_minessale@hotmail.com<mailto:MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<mailto:PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net<http://irc.freenode.net> #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org<mailto:sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888<http://iax:guest@conference.freeswitch.org/888> googletalk:conf+888@conference.freeswitch.org<mailto:googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:213-799-1400
The default config has this in switch.conf.xml <param name="sessions-per-second" value="30"/> /b On Jul 16, 2008, at 9:13 AM, Patrick Grondin wrote:
Just a follow up on the new results I got while testing.
I had to do a couple of tests to figure if my problem was related to my code or the problem anthm figured out.
After a couple of tests, I can say that my problem was mostly related to the way my dialplan was made. The extensions I had added to my dialplan where at the very end of the default dialplan, so for every call I made to these extensions, it had to go through the whole default dialplan which slowed down things.
After putting my test extensions at the top of the file, I can gracefully achieve 30 CPS and I can see that I’m CPU bound to get more performance.
participants (3)
-
Anthony Minessale
-
Brian West
-
Patrick Grondin