[Spce-user] Kamailio Crashes

Fri Sep 11 06:36:20 EDT 2015

Server has 8GB of memory

root at sip:~# free mem
             total       used       free     shared    buffers     cached
Mem:       8197516    5962320    2235196          0     802356    2637092
-/+ buffers/cache:    2522872    5674644
Swap:       901116          0     901116

Yes, the memory was increased after installation.

Greg

Greg Lipschitz | Director | The Summit Group
E: Greg at thesummitgroup.com.au  W: www.thesummitgroup.com.au
The Summit Group (Australia) Pty Ltd | P: 1300 049 749 | Level 1, 39 Railway Road, Blackburn  VIC  3130
The Summit Group (USA) LLC | P: 321 216 3844 | Suite 561, 40E Main Street, Newmark  DE 19711
Postal:      P.O. Box 3225, Doncaster East  VIC  3109

________________________________________
From: Spce-user [spce-user-bounces at lists.sipwise.com] on behalf of Daniel Grotti [dgrotti at sipwise.com]
Sent: Friday, 11 September 2015 8:26 PM
To: spce-user at lists.sipwise.com
Subject: Re: [Spce-user] Kamailio Crashes

mmm...
248 of share mem ? which mean you are running  4G of RAM ?
Did you increase the RAM on your system after SPCE installation ?

--
Daniel Grotti
VoIP Engineer

Sipwise GmbH
Europaring F15 | 2345 Brunn am Gebirge, Austria | www.sipwise.com

On 09/11/2015 12:19 PM, Greg Lipschitz wrote:
> Hi Alex,
>
> Thanks for your prompt reply.
>
> The values you mention are;
>
> pkg_mem: 16
> shm_mem: 248
> tcp_children: 8
> udp_children: 8
>
>
>
> Here is the excerpt from config.yml;
>
> ===START===
> kamailio:
>   lb:
>     debug: 'no'
>     extra_sockets:
>       aapt: udp:10.237.x.x:5060
>     max_forwards: 70
>     nattest_exception_ips: ~
>     pkg_mem: 16
>     port: 5060
>     security:
>       dos_ban_enable: 'yes'
>       dos_ban_time: 300
>       dos_reqs_density_per_unit: 50
>       dos_sampling_time_unit: 5
>       dos_whitelisted_ips: 10.255.255.10, 202.10.x.x, 103.247.x.x
>       dos_whitelisted_subnets: ~
>       failed_auth_attempts: 3
>       failed_auth_ban_enable: 'yes'
>       failed_auth_ban_time: 3600
>     shm_mem: 248
>     start: 'yes'
>     strict_routing_safe: 'no'
>     tcp_children: 8
>     tcp_max_connections: 2048
>     tls:
>       enable: 'yes'
>       port: 5061
>       sslcertfile: '/etc/kamailio/kamailio-selfsigned.pem'
>       sslcertkeyfile: '/etc/kamailio/kamailio-selfsigned.key'
>     udp_children: 8
>     use_dns_cache: 'on'
>   proxy:
>     allow_info_method: 'no'
>     allow_peer_relay: 'no'
>     allow_refer_method: 'no'
>     authenticate_bye: 'no'
>     cf_depth_limit: 10
>     children: 8
>     debug: 'no'
>     default_expires: 3600
>     enum_suffix: e164.arpa.
>     filter_100rel_from_supported: 'no'
>     foreign_domain_via_peer: 'no'
>     fritzbox:
>       enable: 'no'
>       prefixes:
>         - 112
>         - 110
>         - 118[0-9]{2}
>     ignore_auth_realm: 'no'
>     keep_original_to: 'no'
>     max_expires: 43200
>     max_gw_lcr: 128
>     max_registrations_per_subscriber: 5
>     min_expires: 60
>     nathelper_dbro: 'no'
>     natping_interval: 30
>     natping_processes: 7
>     nonce_expire: 300
>     pbx:
>       hunt_display_indicator: '[h]'
>     perform_peer_lcr: 0
>     pkg_mem: 32
>     port: 5062
>     presence:
>       enable: 'yes'
>     proxy_lookup: 'no'
>     set_ruri_to_peer_auth_realm: 'no'
>     shm_mem: 248
>     start: 'yes'
>     tcp_children: 8
>     use_enum: 'no'
>     usrloc_dbmode: 1
> ===END====
>
> Thanks
>
> Greg
>
> ________________________________________
> From: Spce-user [spce-user-bounces at lists.sipwise.com] on behalf of Alex Lutay [alutay at sipwise.com]
> Sent: Friday, 11 September 2015 8:11 PM
> To: spce-user at lists.sipwise.com
> Subject: Re: [Spce-user] Kamailio Crashes
>
> Dear Greg,
>
> Are you sure you have enough memory? Error itself seys:
> tls: openssl bug #1491 (crash/mem leaks on low memory) ...
>
> So you need to check config.yml options you have:
>> kamailio:
>>   lb:
>>     pkg_mem: 16
>>     shm_mem: 125
>>     tcp_children: X
>>     udp_children: Y
>
> As Kamailio is trying to allocate all necessary memory on start:
> allocated = children * pkg_mem (Mb) + shm_mem (Mb) + internal kamailio
> threads/children.
>
> So, if free RAM is not enough to start -> crash. It depends on X and Y
> you have. I believe you have default pkg_mem and shm_mem values.
>
> On 09/11/2015 11:55 AM, Greg Lipschitz wrote:
>> We have been experiencing some random crashing. I happened to catch the logs tonight and this is what was there;
>>
>> Sep 11 19:00:33 sip lb[3238]: NOTICE: <core> [main.c:733]: handle_sigs(): Thank you for flying kamailio!!!
>> Sep 11 19:00:34 sip lb[12766]: WARNING: sl [../../modules/tm/tm_load.h:154]: load_tm_api(): Cannot import load_tm function from tm module
>> Sep 11 19:00:34 sip lb[12766]: WARNING: tls [tls_init.c:636]: init_tls_h(): tls: openssl bug #1491 (crash/mem leaks on low memory) workaround enabled (on low memory tls operations will fail preemptively) with free memory thresholds 25690112 and 12845056 bytes
>> Sep 11 19:00:34 sip lb[12766]: WARNING: sl [../../modules/tm/tm_load.h:154]: load_tm_api(): Cannot import load_tm function from tm module
>>
>> Bucketloads of memory available on the box;
>>
>> root at sip:~# free mem
>>              total       used       free     shared    buffers     cached
>> Mem:       8197516    5925268    2272248          0     802356    2639916
>> -/+ buffers/cache:    2482996    5714520
>> Swap:       901116          0     901116
>>
>> Seems to be the SIP load balancer that is crashing - any ideas?   Just a bug?
>>
>> root at sip:~# ngcpcfg version
>> ngcpcfg, version 0.25.2.1+0~mr3.8.2.1
>
> --
> Alexander Lutay
> _______________________________________________
> Spce-user mailing list
> Spce-user at lists.sipwise.com
> https://lists.sipwise.com/listinfo/spce-user
>
>
>
> _______________________________________________
> Spce-user mailing list
> Spce-user at lists.sipwise.com
> https://lists.sipwise.com/listinfo/spce-user
>
_______________________________________________
Spce-user mailing list
Spce-user at lists.sipwise.com
https://lists.sipwise.com/listinfo/spce-user