VoIP Mailing List Archives :: View topic - [asterisk-users] asterisk optimalization

Sponsor: VoiceMeUp - Corporate & Wholesale VoIP Services

VoIP Mailing List Archives
Mailing list archives for the VoIP community

[asterisk-users] asterisk optimalization

VoIP Mailing List Archives Forum Index -> Asterisk Users

View previous topic :: View next topic

Author

Message

cervajs at fpf.slu.cz
Guest

Posted: Wed Jan 23, 2008 11:53 am Post subject: [asterisk-users] asterisk optimalization

hi, i'm testing asterisk 1.4/1.2 in the following scenario centos5/cpu quad xeon E5335 2.0Ghz - test clients behind nat - 1500+ testing instances - reregister option from 1min to 1hour - qualify set to 5000 top shows over 100% cpu. cpu cores sometimes go to 95% with htop i see ~16threads but only one child have ~95% cpu (how i can get info about that thread? what he is doing?) what is major bottleneck? qualify imho not. i'm tried set qualify=no, does not help SIP REGISTER packets? this problem persist if no calls are active after restart cpu usage slowly increase. after a ~hour is about 100% which optimalizations do you recommend for ~1500 peers scenario? (behind nat, reregistrations) --------------------------------------- Marek Cervenka =======================================

goksie at gmail.com
Guest

Posted: Thu Jan 24, 2008 1:32 am Post subject: [asterisk-users] asterisk optimalization

ram wrote:

Quote:

http://www.transnexus.com/White%20Papers/asterisk_V1-4-11_performance.htm

check this link may help you

ram

On Jan 23, 2008 10:23 PM, marek cervenka <cervajs at fpf.slu.cz
<mailto:cervajs at fpf.slu.cz>> wrote:

hi,

i'm testing asterisk 1.4/1.2 in the following scenario
centos5/cpu quad xeon E5335 2.0Ghz
- test clients behind nat
- 1500+ testing instances - reregister option from 1min to 1hour
- qualify set to 5000

top shows over 100% cpu. cpu cores sometimes go to 95%
with htop i see ~16threads but only one child have ~95% cpu
(how i can get info about that thread? what he is doing?)

what is major bottleneck? qualify imho not. i'm tried set
qualify=no, does not help
SIP REGISTER packets?

this problem persist if no calls are active
after restart cpu usage slowly increase. after a ~hour is about 100%

which optimalizations do you recommend for ~1500 peers scenario? (behind
nat, reregistrations)

---------------------------------------
Marek Cervenka
=======================================

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com
<http://www.api-digital.com/> --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users
<http://lists.digium.com/mailman/listinfo/asterisk-users>

------------------------------------------------------------------------

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users

gopal.krishnan at peop...
Guest

Posted: Thu Jan 24, 2008 2:50 am Post subject: [asterisk-users] asterisk optimalization

Hi,

Dell is not a recomeded server for linux. Its only compatible with
windows.

On Jan 24, 2008 12:02 PM, Goke Aruna <goksie at gmail.com> wrote:

Quote:

ram wrote:
http://www.transnexus.com/White%20Papers/asterisk_V1-4-11_performance.htm

Quote:

check this link may help you

ram

On Jan 23, 2008 10:23 PM, marek cervenka <cervajs at fpf.slu.cz
<mailto:cervajs at fpf.slu.cz>> wrote:

hi,

i'm testing asterisk 1.4/1.2 in the following scenario
centos5/cpu quad xeon E5335 2.0Ghz
- test clients behind nat
- 1500+ testing instances - reregister option from 1min to 1hour
- qualify set to 5000

top shows over 100% cpu. cpu cores sometimes go to 95%
with htop i see ~16threads but only one child have ~95% cpu
(how i can get info about that thread? what he is doing?)

what is major bottleneck? qualify imho not. i'm tried set
qualify=no, does not help
SIP REGISTER packets?

this problem persist if no calls are active
after restart cpu usage slowly increase. after a ~hour is about 100%

which optimalizations do you recommend for ~1500 peers scenario?

(behind

Quote:

nat, reregistrations)

---------------------------------------
Marek Cervenka
=======================================

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com
<http://www.api-digital.com/> --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users
<http://lists.digium.com/mailman/listinfo/asterisk-users>

------------------------------------------------------------------------

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users

--
Thank you with regards,
Gopal,
PeopleTech Systems Private Limited
www.peopletech.co.in
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20080124/6e01bc7c/attachment.htm

tzafrir.cohen at xorco...
Guest

Posted: Thu Jan 24, 2008 3:02 am Post subject: [asterisk-users] asterisk optimalization

On Thu, Jan 24, 2008 at 01:20:59PM +0530, Gopal krishnan wrote:

Quote:

Hi,

Dell is not a recomeded server for linux. Its only compatible with
windows.

And I suppose you have checked this. And specifically paid a short visit
to Dell's site before.

http://linux.dell.com
http://linux.dell.com/distributions.shtml

--
Tzafrir Cohen
icq#16849755 jabber:tzafrir.cohen at xorcom.com
+972-50-7952406 mailto:tzafrir.cohen at xorcom.com
http://www.xorcom.com iax:guest at local.xorcom.com/tzafrir

benny+usenet at amorse...
Guest

Posted: Thu Jan 24, 2008 5:43 am Post subject: [asterisk-users] asterisk optimalization

marek cervenka <cervajs at fpf.slu.cz> writes:

Quote:

hi,

i'm testing asterisk 1.4/1.2 in the following scenario
centos5/cpu quad xeon E5335 2.0Ghz
- test clients behind nat
- 1500+ testing instances - reregister option from 1min to 1hour
- qualify set to 5000

top shows over 100% cpu. cpu cores sometimes go to 95%
with htop i see ~16threads but only one child have ~95% cpu
(how i can get info about that thread? what he is doing?)

oprofile can probably tell you. It can be a bit difficult to get
all the debugging information into the right places so oprofile works,
but it's very helpful.
/Benny

daniel.guthrie at appl...
Guest

Posted: Thu Jan 24, 2008 9:55 am Post subject: [asterisk-users] asterisk optimalization

That's funny. I seem to remember installed Deb/* on a Poweredge 2950.

.must be slowly losing my mind. Another side effect of using Asterisk?
Dementia?

From: asterisk-users-bounces at lists.digium.com
[mailto:asterisk-users-bounces at lists.digium.com] On Behalf Of Gopal krishnan
Sent: Thursday, January 24, 2008 2:51 AM
To: goksie at gmail.com; Asterisk Users Mailing List - Non-Commercial
Discussion
Subject: Re: [asterisk-users] asterisk optimalization

Hi,

Dell is not a recomeded server for linux. Its only compatible with
windows.

On Jan 24, 2008 12:02 PM, Goke Aruna <goksie at gmail.com> wrote:

ram wrote:

Quote:

http://www.transnexus.com/White%20Papers/asterisk_V1-4-11_performance.htm

check this link may help you

ram

On Jan 23, 2008 10:23 PM, marek cervenka < cervajs at fpf.slu.cz

<mailto:cervajs at fpf.slu.cz>

Quote:

<mailto:cervajs at fpf.slu.cz>> wrote:

hi,

i'm testing asterisk 1.4/1.2 in the following scenario
centos5/cpu quad xeon E5335 2.0Ghz
- test clients behind nat
- 1500+ testing instances - reregister option from 1min to 1hour
- qualify set to 5000

top shows over 100% cpu. cpu cores sometimes go to 95%
with htop i see ~16threads but only one child have ~95% cpu
(how i can get info about that thread? what he is doing?)

what is major bottleneck? qualify imho not. i'm tried set
qualify=no, does not help
SIP REGISTER packets?

this problem persist if no calls are active
after restart cpu usage slowly increase. after a ~hour is about 100%

which optimalizations do you recommend for ~1500 peers scenario?

(behind

Quote:

------------------------------------------------------------------------

Quote:

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
<http://lists.digium.com/mailman/listinfo/asterisk-users>

http://lists.digium.com/mailman/listinfo/asterisk-users

http://www.transnexus.com/White%20Papers/asterisk_V1-4-11_performance.htm

That result is suprising! but i have DELL 2950 with 2 X 3.0GHz CPU on
6GB ram, equiped with 8e1 link (2 sangoma A104D) running FC5. I
installed chan_ss7-1.0 with asterisk-1.2.25 doing transcoding, and each
time calls get to 120+ the cpu is fully utilized.

the calls come from sip to the ss7 link.

can someone advice me on what I can do to improve the performance.
goksie
NB. I felt we re talking on the same topic thats why i added my own
experience.

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com
<http://www.api-digital.com> --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users
<http://lists.digium.com/mailman/listinfo/asterisk-users>

--
Thank you with regards,
Gopal,
PeopleTech Systems Private Limited
www.peopletech.co.in

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20080124/ed63939e/attachment-0001.htm

asteriskal at gmail.com
Guest

Posted: Thu Jan 24, 2008 9:49 pm Post subject: [asterisk-users] asterisk optimalization

I have been using Dell servers and have no issues with linux, in fact when i
implemented my last install with their top of the line server (dual xeon
quad core and SAS drives on Perc 6i) i was amazed how smoothly it went
trough.
Beside that i like their open manage, it runs nice on linux and its a handy
tool on remote location servers.
On Jan 24, 2008 7:55 AM, Daniel Guthrie <
daniel.guthrie at applewoodcommunications.com> wrote:

Quote:

That's funny? I seem to remember installed Deb/* on a Poweredge 2950?

?must be slowly losing my mind. Another side effect of using Asterisk?
Dementia?

*From:* asterisk-users-bounces at lists.digium.com [mailto:
asterisk-users-bounces at lists.digium.com] *On Behalf Of *Gopal krishnan
*Sent:* Thursday, January 24, 2008 2:51 AM
*To:* goksie at gmail.com; Asterisk Users Mailing List - Non-Commercial
Discussion
*Subject:* Re: [asterisk-users] asterisk optimalization

Hi,

Dell is not a recomeded server for linux. Its only compatible with
windows.

On Jan 24, 2008 12:02 PM, Goke Aruna <goksie at gmail.com> wrote:

ram wrote:
http://www.transnexus.com/White%20Papers/asterisk_V1-4-11_performance.htm

Quote:

check this link may help you

ram

On Jan 23, 2008 10:23 PM, marek cervenka < cervajs at fpf.slu.cz

Quote:

(behind

Quote:

------------------------------------------------------------------------

Quote:

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users

http://www.transnexus.com/White%20Papers/asterisk_V1-4-11_performance.htm

That result is suprising! but i have DELL 2950 with 2 X 3.0GHz CPU on
6GB ram, equiped with 8e1 link (2 sangoma A104D) running FC5. I
installed chan_ss7-1.0 with asterisk-1.2.25 doing transcoding, and each
time calls get to 120+ the cpu is fully utilized.

the calls come from sip to the ss7 link.

can someone advice me on what I can do to improve the performance.

goksie
NB. I felt we re talking on the same topic thats why i added my own
experience.

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users

--
Thank you with regards,
Gopal,
PeopleTech Systems Private Limited
www.peopletech.co.in

_______________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20080124/76e92df4/attachment-0001.htm

cervajs at fpf.slu.cz
Guest

Posted: Fri Jan 25, 2008 12:05 am Post subject: [asterisk-users] asterisk optimalization

Quote:

marek cervenka <cervajs at fpf.slu.cz> writes:

Quote:

oprofile can probably tell you. It can be a bit difficult to get
all the debugging information into the right places so oprofile works,
but it's very helpful.

this is strace -p ppid_of_problematic_thread

can you look if you see any anomalies?

--cut--
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 14
connect(14, {sa_family=AF_INET, sin_port=htons(50195),
sin_addr=inet_addr("filtered")}, 16) = 0
getsockname(14, {sa_family=AF_INET, sin_port=htons(32777),
sin_addr=inet_addr("filtered")}, [16]) = 0
close(14) = 0
time(NULL) = 1201236754
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
gettimeofday({1201236754, 532554}, NULL) = 0
sendto(11, "OPTIONS sip:filtered at 10.72.20.4"..., 497, 0,
{sa_family=AF_INET, sin_port=htons(5060), sin_addr=inet_addr("filtered")},
16) = 497
gettimeofday({1201236754, 532798}, NULL) = 0
gettimeofday({1201236754, 532909}, NULL) = 0
gettimeofday({1201236754, 533010}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 533903}, NULL) = 0
poll([{fd=11, events=POLLIN}], 1, 17) = 0
gettimeofday({1201236754, 551423}, NULL) = 0
gettimeofday({1201236754, 551535}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 551994}, NULL) = 0
poll([{fd=11, events=POLLIN, revents=POLLIN}], 1, 20) = 1
recvfrom(11, "SIP/2.0 200 OK\r\nTo: <sip:filtered"..., 4095, 0,
{sa_family=AF_INET, sin_port=htons(5060), sin_addr=inet_addr("filtered")},
[16]) = 422
gettimeofday({1201236754, 557006}, NULL) = 0
gettimeofday({1201236754, 557065}, NULL) = 0
gettimeofday({1201236754, 557397}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 557794}, NULL) = 0
poll([{fd=11, events=POLLIN}], 1, 14) = 0
gettimeofday({1201236754, 571604}, NULL) = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 14
connect(14, {sa_family=AF_INET, sin_port=htons(50195),
sin_addr=inet_addr("filtered")}, 16) = 0
getsockname(14, {sa_family=AF_INET, sin_port=htons(32777),
sin_addr=inet_addr("filtered")}, [16]) = 0
close(14) = 0
time(NULL) = 1201236754
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
gettimeofday({1201236754, 572328}, NULL) = 0
sendto(11, "OPTIONS sip:filtered at 10.0.0.4 S"..., 481, 0,
{sa_family=AF_INET, sin_port=htons(5060), sin_addr=inet_addr("filtered")},
16) = 481
gettimeofday({1201236754, 572462}, NULL) = 0
gettimeofday({1201236754, 572498}, NULL) = 0
gettimeofday({1201236754, 572566}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 572859}, NULL) = 0
poll([{fd=11, events=POLLIN}], 1, 1) = 0
gettimeofday({1201236754, 573604}, NULL) = 0
gettimeofday({1201236754, 573651}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 573872}, NULL) = 0
poll([{fd=11, events=POLLIN}], 1, 5) = 0
gettimeofday({1201236754, 578602}, NULL) = 0
gettimeofday({1201236754, 578652}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 578863}, NULL) = 0
poll([{fd=11, events=POLLIN, revents=POLLIN}], 1, 11) = 1
recvfrom(11, "\0\0\0\0", 4095, 0, {sa_family=AF_INET,
sin_port=htons(5060), sin_addr=inet_addr("filtered")}, [16]) = 4
gettimeofday({1201236754, 587424}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 587639}, NULL) = 0
poll([{fd=11, events=POLLIN}], 1, 2) = 0
gettimeofday({1201236754, 589599}, NULL) = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 14
connect(14, {sa_family=AF_INET, sin_port=htons(50195),
sin_addr=inet_addr("filtered")}, 16) = 0
getsockname(14, {sa_family=AF_INET, sin_port=htons(32777),
sin_addr=inet_addr("filtered")}, [16]) = 0
close(14) = 0
time(NULL) = 1201236754
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
gettimeofday({1201236754, 590247}, NULL) = 0
sendto(11, "OPTIONS sip:filtered at 10.0.222.2"..., 499, 0,
{sa_family=AF_INET, sin_port=htons(5060), sin_addr=inet_addr("filtered")},
16) = 499
gettimeofday({1201236754, 590382}, NULL) = 0
gettimeofday({1201236754, 590424}, NULL) = 0
gettimeofday({1201236754, 590479}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 590742}, NULL) = 0
poll([{fd=11, events=POLLIN}], 1, 3) = 0
gettimeofday({1201236754, 593600}, NULL) = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 14
connect(14, {sa_family=AF_INET, sin_port=htons(50195),
sin_addr=inet_addr("filtered")}, 16) = 0
getsockname(14, {sa_family=AF_INET, sin_port=htons(32777),
sin_addr=inet_addr("filtered")}, [16]) = 0
close(14) = 0
time(NULL) = 1201236754
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=806, ...}) = 0
gettimeofday({1201236754, 594242}, NULL) = 0
sendto(11, "OPTIONS sip:filtered at 192.168.1."..., 491, 0,
{sa_family=AF_INET, sin_port=htons(5060), sin_addr=inet_addr("filtered")},
16) = 491
gettimeofday({1201236754, 594372}, NULL) = 0
gettimeofday({1201236754, 594415}, NULL) = 0
gettimeofday({1201236754, 594465}, NULL) = 0
time(NULL) = 1201236754
time(NULL) = 1201236754
gettimeofday({1201236754, 594731}, NULL) = 0
--cut--
---------------------------------------
Marek Cervenka
=======================================

cervajs at fpf.slu.cz
Guest

Posted: Fri Jan 25, 2008 12:31 am Post subject: [asterisk-users] asterisk optimalization

Quote:

marek cervenka <cervajs at fpf.slu.cz> writes:

Quote:

oprofile can probably tell you. It can be a bit difficult to get
all the debugging information into the right places so oprofile works,
but it's very helpful.

this is strace -p ppid_of_problematic_thread

can you look if you see any anomalies?

one more info
with
iptables -A INPUT -p udp -s 0/0 -d 0/0 --dport 5060 -j REJECT
cpu usage goes to 0-4% -> this problem is not some asterisk cpu deadlock
but problem in process incoming packets

---------------------------------------
Marek Cervenka
=======================================

Display posts from previous:

	VoIP Mailing List Archives Forum Index -> Asterisk Users	All times are GMT - 5 Hours
Page 1 of 1

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum