Re: [OpenSIPS-Users] Opensips stops responding to TLS

[L S]

Hi Răzvan,
Just saw your response.

> Does it stop to any TLS operation, even for new ones?
All TLS operations stop, no TLS traffic on Wireshark.

We are using openssl 1.1.1q on Centos 7.
We wanted to try wolfssl, but we had issues installing it on Centos 7. If
it is more stable than openssl, we can give it another shot.

Thanks,
Matt

On Thu, Feb 2, 2023, 6:18 AM Răzvan Crainea  wrote:

> Hello!
>
> Does it stop to any TLS operation, even for new ones? What TLS lib are
> you using, openssl or wolfssl?
> Are there any errors in the logs related to TLS?
>
> Best regards,
>
> Răzvan Crainea
> OpenSIPS Core Developer
> http://www.opensips-solutions.com
>
> On 12/30/22 16:00, L S wrote:
> > One more thing:
> >
> > log_level=4
> > open_files_limit=32768
> >
> > At the time Opensips stops responding to TLS, it seems like it stops
> > writing to log file too even though it continues handling the non-TLS
> SIP.
> >
> > Thanks.
> >
> > On Thu, Dec 29, 2022, 5:51 PM L S  > > wrote:
> >
> > Just wanted to add the traffic between the client and Opensips
> > below. It seems Opensips keeps on sending RESET.
> >
> > We have the tcp_max_connections at default value. That value (2048)
> > works fine in 1.11.5.
> >
> > Thanks.
> >
> > client opensipsSSL142Client Hello
> > client opensipsSSL142[TCP Retransmission] Client Hello
> > opensipsclient TCP54sips > 5071 [RST] Seq=1 Win=0 Len=0
> > client opensipsSSL142[TCP Retransmission] Client Hello
> > opensipsclient TCP54sips > 5064 [RST] Seq=1 Win=0 Len=0
> > client opensipsTCP74[TCP Port numbers reused] 5071 > sips [SYN]
> > Seq=0 Win=8192 Len=0 MSS=1460 WS=1
> > opensipsclient TCP54sips > 5071 [RST, ACK] Seq=1 Ack=1 Win=0 Len=0
> > client opensipsTCP74[TCP Port numbers reused] 5064 > sips [SYN]
> > Seq=0 Win=8192 Len=0 MSS=1460 WS=1
> > opensipsclient TCP54sips > 5064 [RST, ACK] Seq=1 Ack=1 Win=0 Len=0
> > client opensipsTCP74[TCP Port numbers reused] 5080 > sips [SYN]
> > Seq=0 Win=8192 Len=0 MSS=1460 WS=1
> >
> > On Thu, Dec 29, 2022, 9:27 AM L S  > > wrote:
> >
> > Hi,
> >
> > We are in the process of migrating from 1.11.5 tls to 3.2.9, and
> > we are running into an issue with TLS.
> >
> > Opensips stops handling TLS within a few minutes after it is
> > started; e.g. stops responding to Client Hellos. There is no
> > more outgoing TLS traffic from the Opensips server. When we
> > restart Opensips, it goes back to normal for a while, then stops
> > responding to TLS requests again.
> >
> > I don't see any errors in logs.
> > The server runs Centos 7,  openssl 1.1.1q.
> >
> > 1.11.5 works fine.
> >
> > Can this be a memory issue? We use S_memory 512 and P_memory 8.
> > Opensips 1.11.5 works fine with the same settings. TCP
> > parameters have their default values.
> >
> > How can we debug this? Any suggestions would be appreciated.
> >
> > Thanks,
> > Matt
> >
> >
> > ___
> > Users mailing list
> > Users@lists.opensips.org
> > http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
> ___
> Users mailing list
> Users@lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] Opensips stops responding to TLS

[Răzvan Crainea]

Hello!

Does it stop to any TLS operation, even for new ones? What TLS lib are 
you using, openssl or wolfssl?

Are there any errors in the logs related to TLS?

Best regards,

Răzvan Crainea
OpenSIPS Core Developer
http://www.opensips-solutions.com

On 12/30/22 16:00, L S wrote:

One more thing:

log_level=4
open_files_limit=32768

At the time Opensips stops responding to TLS, it seems like it stops 
writing to log file too even though it continues handling the non-TLS SIP.


Thanks.

On Thu, Dec 29, 2022, 5:51 PM L S > wrote:


Just wanted to add the traffic between the client and Opensips
below. It seems Opensips keeps on sending RESET.

We have the tcp_max_connections at default value. That value (2048)
works fine in 1.11.5.

Thanks.

client opensipsSSL142Client Hello
client opensipsSSL142[TCP Retransmission] Client Hello
opensipsclient TCP54sips > 5071 [RST] Seq=1 Win=0 Len=0
client opensipsSSL142[TCP Retransmission] Client Hello
opensipsclient TCP54sips > 5064 [RST] Seq=1 Win=0 Len=0
client opensipsTCP74[TCP Port numbers reused] 5071 > sips [SYN]
Seq=0 Win=8192 Len=0 MSS=1460 WS=1
opensipsclient TCP54sips > 5071 [RST, ACK] Seq=1 Ack=1 Win=0 Len=0
client opensipsTCP74[TCP Port numbers reused] 5064 > sips [SYN]
Seq=0 Win=8192 Len=0 MSS=1460 WS=1
opensipsclient TCP54sips > 5064 [RST, ACK] Seq=1 Ack=1 Win=0 Len=0
client opensipsTCP74[TCP Port numbers reused] 5080 > sips [SYN]
Seq=0 Win=8192 Len=0 MSS=1460 WS=1

On Thu, Dec 29, 2022, 9:27 AM L S mailto:efes99...@gmail.com>> wrote:

Hi,

We are in the process of migrating from 1.11.5 tls to 3.2.9, and
we are running into an issue with TLS.

Opensips stops handling TLS within a few minutes after it is
started; e.g. stops responding to Client Hellos. There is no
more outgoing TLS traffic from the Opensips server. When we
restart Opensips, it goes back to normal for a while, then stops
responding to TLS requests again.

I don't see any errors in logs.
The server runs Centos 7,  openssl 1.1.1q.

1.11.5 works fine.

Can this be a memory issue? We use S_memory 512 and P_memory 8.
Opensips 1.11.5 works fine with the same settings. TCP
parameters have their default values.

How can we debug this? Any suggestions would be appreciated.

Thanks,
Matt


___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users


___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] Opensips stops responding to TLS

[L S]

One more thing:

log_level=4
open_files_limit=32768

At the time Opensips stops responding to TLS, it seems like it stops
writing to log file too even though it continues handling the non-TLS SIP.

Thanks.

On Thu, Dec 29, 2022, 5:51 PM L S  wrote:

> Just wanted to add the traffic between the client and Opensips below. It
> seems Opensips keeps on sending RESET.
>
> We have the tcp_max_connections at default value. That value (2048) works
> fine in 1.11.5.
>
> Thanks.
>
> clientopensips SSL 142 Client Hello
> clientopensips SSL 142 [TCP Retransmission] Client Hello
> opensips clientTCP 54 sips > 5071 [RST] Seq=1 Win=0 Len=0
> clientopensips SSL 142 [TCP Retransmission] Client Hello
> opensips clientTCP 54 sips > 5064 [RST] Seq=1 Win=0 Len=0
> clientopensips TCP 74 [TCP Port numbers reused] 5071 > sips [SYN]
> Seq=0 Win=8192 Len=0 MSS=1460 WS=1
> opensips clientTCP 54 sips > 5071 [RST, ACK] Seq=1 Ack=1 Win=0 Len=0
> clientopensips TCP 74 [TCP Port numbers reused] 5064 > sips [SYN]
> Seq=0 Win=8192 Len=0 MSS=1460 WS=1
> opensips clientTCP 54 sips > 5064 [RST, ACK] Seq=1 Ack=1 Win=0 Len=0
> clientopensips TCP 74 [TCP Port numbers reused] 5080 > sips [SYN]
> Seq=0 Win=8192 Len=0 MSS=1460 WS=1
>
> On Thu, Dec 29, 2022, 9:27 AM L S  wrote:
>
>> Hi,
>>
>> We are in the process of migrating from 1.11.5 tls to 3.2.9, and we are
>> running into an issue with TLS.
>>
>> Opensips stops handling TLS within a few minutes after it is started;
>> e.g. stops responding to Client Hellos. There is no more outgoing TLS
>> traffic from the Opensips server. When we restart Opensips, it goes back to
>> normal for a while, then stops responding to TLS requests again.
>>
>> I don't see any errors in logs.
>> The server runs Centos 7,  openssl 1.1.1q.
>>
>> 1.11.5 works fine.
>>
>> Can this be a memory issue? We use S_memory 512 and P_memory 8. Opensips
>> 1.11.5 works fine with the same settings. TCP parameters have their default
>> values.
>>
>> How can we debug this? Any suggestions would be appreciated.
>>
>> Thanks,
>> Matt
>>
>
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] Opensips stops responding to TLS

[L S]

Just wanted to add the traffic between the client and Opensips below. It
seems Opensips keeps on sending RESET.

We have the tcp_max_connections at default value. That value (2048) works
fine in 1.11.5.

Thanks.

clientopensips SSL 142 Client Hello
clientopensips SSL 142 [TCP Retransmission] Client Hello
opensips clientTCP 54 sips > 5071 [RST] Seq=1 Win=0 Len=0
clientopensips SSL 142 [TCP Retransmission] Client Hello
opensips clientTCP 54 sips > 5064 [RST] Seq=1 Win=0 Len=0
clientopensips TCP 74 [TCP Port numbers reused] 5071 > sips [SYN] Seq=0
Win=8192 Len=0 MSS=1460 WS=1
opensips clientTCP 54 sips > 5071 [RST, ACK] Seq=1 Ack=1 Win=0 Len=0
clientopensips TCP 74 [TCP Port numbers reused] 5064 > sips [SYN] Seq=0
Win=8192 Len=0 MSS=1460 WS=1
opensips clientTCP 54 sips > 5064 [RST, ACK] Seq=1 Ack=1 Win=0 Len=0
clientopensips TCP 74 [TCP Port numbers reused] 5080 > sips [SYN] Seq=0
Win=8192 Len=0 MSS=1460 WS=1

On Thu, Dec 29, 2022, 9:27 AM L S  wrote:

> Hi,
>
> We are in the process of migrating from 1.11.5 tls to 3.2.9, and we are
> running into an issue with TLS.
>
> Opensips stops handling TLS within a few minutes after it is started; e.g.
> stops responding to Client Hellos. There is no more outgoing TLS traffic
> from the Opensips server. When we restart Opensips, it goes back to normal
> for a while, then stops responding to TLS requests again.
>
> I don't see any errors in logs.
> The server runs Centos 7,  openssl 1.1.1q.
>
> 1.11.5 works fine.
>
> Can this be a memory issue? We use S_memory 512 and P_memory 8. Opensips
> 1.11.5 works fine with the same settings. TCP parameters have their default
> values.
>
> How can we debug this? Any suggestions would be appreciated.
>
> Thanks,
> Matt
>
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

[OpenSIPS-Users] Opensips stops responding to TLS

[L S]

Hi,

We are in the process of migrating from 1.11.5 tls to 3.2.9, and we are
running into an issue with TLS.

Opensips stops handling TLS within a few minutes after it is started; e.g.
stops responding to Client Hellos. There is no more outgoing TLS traffic
from the Opensips server. When we restart Opensips, it goes back to normal
for a while, then stops responding to TLS requests again.

I don't see any errors in logs.
The server runs Centos 7,  openssl 1.1.1q.

1.11.5 works fine.

Can this be a memory issue? We use S_memory 512 and P_memory 8. Opensips
1.11.5 works fine with the same settings. TCP parameters have their default
values.

How can we debug this? Any suggestions would be appreciated.

Thanks,
Matt
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] Opensips stops responding

[Răzvan Crainea]

Hi, Schneur!

If this happens again, can you please run an `opensipsctl trap` command 
- this could help us understand why it was not handling traffic.
BTW: could it be possible that the keepalived script also removed the IP 
from the interface, and OpenSIPS was no longer getting any traffic 
whatsoever?


Best regards,
Răzvan

On 7/12/19 1:38 PM, Schneur Rosenberg wrote:
Another update, seems like my assumption that some requests were handeld 
was wrong,  the keepalive script had a bug and therefore didn't 
failover, I now believe OpenSIPs ignored all requests after the initial 
DB error, this seems like a bug in OpenSIPs it shouldnt just stop 
responding to requests because of one DB attempt that went wrong.


On Thu, Jul 11, 2019, 12:39 PM S. Rosenberg > wrote:


I want to add that a restart of OpenSIPs fixed the problem every time, I
didnt restart the MySQL server.



--
Sent from:

http://opensips-open-sip-server.1449251.n2.nabble.com/OpenSIPS-Users-f1449235.html

___
Users mailing list
Users@lists.opensips.org 
http://lists.opensips.org/cgi-bin/mailman/listinfo/users


___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users



--
Răzvan Crainea
OpenSIPS Core Developer
  http://www.opensips-solutions.com

___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] Opensips stops responding

[Schneur Rosenberg]

Another update, seems like my assumption that some requests were handeld
was wrong,  the keepalive script had a bug and therefore didn't failover, I
now believe OpenSIPs ignored all requests after the initial DB error, this
seems like a bug in OpenSIPs it shouldnt just stop responding to requests
because of one DB attempt that went wrong.

On Thu, Jul 11, 2019, 12:39 PM S. Rosenberg 
wrote:

> I want to add that a restart of OpenSIPs fixed the problem every time, I
> didnt restart the MySQL server.
>
>
>
> --
> Sent from:
> http://opensips-open-sip-server.1449251.n2.nabble.com/OpenSIPS-Users-f1449235.html
>
> ___
> Users mailing list
> Users@lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] Opensips stops responding

[S. Rosenberg]

I want to add that a restart of OpenSIPs fixed the problem every time, I
didnt restart the MySQL server.



--
Sent from: 
http://opensips-open-sip-server.1449251.n2.nabble.com/OpenSIPS-Users-f1449235.html

___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

[OpenSIPS-Users] Opensips stops responding

[Schneur Rosenberg]

Hi, I've had recently a few times that OpenSIPSs didnt crash but it
stopped responding to certain requests, perhaps INVITES that requeired
DB access, but I havent verified it becuase I always wanted to bring
it back up ASAP because it seems like it did respond to my requests
that my keepalived VRRP script which sends a notify to OpenSIPs and
OpenSIPS does a DB lookup and only on success it returns a true thats
how I make sure that MYSQL is working in addition to  OpenSIPs, but
opensips seems to have responded because it did not failover, and
thats why I didnt have time to investigate further to which requests
it responds and to which it didnt.

Here are the syslogs from the time of the crash, maybe it can shed
some light to the issue, the query in there that failed is the query
from keepalived, I assume it failed only once and then did respond,
otherwise the IP would of failed over to the backup, but its only an
assumption.

Jul 10 20:13:13 sipsvr1 /sbin/opensips[29446]:
INFO:db_mysql:switch_state_to_disconnected: disconnect event for
0x7f2ba43e7600
Jul 10 20:13:13 sipsvr1 /sbin/opensips[29446]:
INFO:db_mysql:reset_all_statements: resetting all statements on
connection: (0x7f2ba44062a0) 0x7f2ba43e7600
Jul 10 20:13:13 sipsvr1 /sbin/opensips[29446]:
INFO:db_mysql:connect_with_retry: re-connected successful for
0x7f2ba43e7600
Jul 10 20:13:13 sipsvr1 /sbin/opensips[29446]:
INFO:db_mysql:db_mysql_do_prepared_query: reconnected to mysql server
-> re-init the statement
Jul 10 20:13:13 sipsvr1 /sbin/opensips[29486]:
WARNING:core:handle_timer_job: utimer job  has a 3 us
delay in execution
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29400]:
INFO:clusterer:do_action_trans_2: Ping reply not received, node [2] is
down
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29400]:
INFO:clusterer:do_action_trans_2: Ping reply not received, node [2] is
down
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29478]:
INFO:clusterer:handle_internal_msg: Node [2] is UP
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29478]:
INFO:clusterer:handle_internal_msg: Node [2] is UP
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29422]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29425.374568.2]
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29419]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29431.1000819.1]
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29428]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29411.161103.9]
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29446]:
CRITICAL:db_mysql:wrapper_single_mysql_stmt_execute: driver error
(1180): Got error 35 "Resource deadlock avoided" during COMMIT
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29446]:
ERROR:usrloc:db_update_ucontact: updating database failed
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29446]: ERROR:usrloc:wb_timer:
updating contact in db failed
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29388]:
CRITICAL:db_mysql:wrapper_single_mysql_real_query: driver error
(1047): WSREP has not yet prepared node for application use
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29388]:
ERROR:core:db_do_raw_query: error while submitting query
Jul 10 20:13:15 sipsvr1 /sbin/opensips[29388]:
ERROR:avpops:db_query_avp: raw_query failed: db0(usr_preferences)
select domain from domain where domain='...
Jul 10 20:13:15 sipsvr1 Keepalived_vrrp[514]: pid 2092 exited with status 1
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29428]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29422.274079.61]
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29429]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29421.236751.61]
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29425]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29419.224887.28]
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29422]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29413.165448.9]
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29419]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29429.711965.1]
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29411]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29431.1000786.2]
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29428]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29411.161104.10]
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29421]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29419.224888.10]
Jul 10 20:13:16 sipsvr1 /sbin/opensips[29433]:
INFO:presence:update_presentity: *** found in db but not in htable
[a.1561468443.29421.236752.28]

___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] opensips - stops responding to some SIP messages

[Alexey Kazantsev via Users]

Have you tried to trace the script?
OpenSIPS has tools to detect the time of execution
of script parts.

Check this please:
https://blog.opensips.org/2017/09/20/troubleshooting-opensips-script/

---
BR, Alexey
http://alexeyka.zantsev.com/
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

[OpenSIPS-Users] opensips - stops responding to some SIP messages

[Maxim V. Maximov]

Thank you all for the answers, the problem has not yet been solved

>>did you take network traces at the time of the fault, does the traffic
get to the server?

Yes, we checked. check when removing the dump on the network interface, we
see that the request goes to our server

RX errors:0 dropped:0 overruns:0 frame:0
TX errors:0 dropped:0 overruns:0 carrier:0


opensipsctl fifo log_level

Process::  PID=2998 Log level=3 Type=attendant
Process::  PID=3009 Log level=3 Type=event-route handler
Process::  PID=3010 Log level=3 Type=RTPP timeout receiver
Process::  PID=3011 Log level=3 Type=MI Datagram
Process::  PID=3012 Log level=3 Type=MI FIFO
Process::  PID=3013 Log level=3 Type=time_keeper
Process::  PID=3014 Log level=3 Type=timer
Process::  PID=3017 Log level=3 Type=SIP receiver udp:x.x.x.x:5060
Process::  PID=3019 Log level=3 Type=SIP receiver udp:x.x.x.x0:5060
Process::  PID=3021 Log level=3 Type=SIP receiver udp:x.x.x.x:5060
Process::  PID=3023 Log level=3 Type=SIP receiver udp:x.x.x.x:5060
Process::  PID=3025 Log level=3 Type=SIP receiver udp:x.x.x.x:5060
Process::  PID=3028 Log level=3 Type=SIP receiver udp:x.x.x.x:5060
Process::  PID=3031 Log level=3 Type=SIP receiver udp:x.x.x.x:5060
Process::  PID=3032 Log level=3 Type=SIP receiver udp:x.x.x.x:5060
Process::  PID=3033 Log level=3 Type=SIP receiver udp:x.x.x.x:5068
Process::  PID=3034 Log level=3 Type=SIP receiver udp:x.x.x.x:5068
Process::  PID=3035 Log level=3 Type=SIP receiver udp:x.x.x.x:5068
Process::  PID=3036 Log level=3 Type=SIP receiver udp:x.x.x.x:5068
Process::  PID=3037 Log level=3 Type=SIP receiver udp:x.x.x.x:5068
Process::  PID=3038 Log level=3 Type=SIP receiver udp:x.x.x.x:5068
Process::  PID=3039 Log level=3 Type=SIP receiver udp:x.x.x.x:5068
Process::  PID=3040 Log level=3 Type=SIP receiver udp:x.x.x.x:5068
Process::  PID=3041 Log level=3 Type=Timer handler


opensipsctl fifo get_statistics shmem:
shmem:total_size:: 6316621824
shmem:used_size:: 248015312
shmem:real_used_size:: 663246088
shmem:max_used_size:: 1204678592
shmem:free_size:: 5653375736
shmem:fragments:: 4312144

>>Also, is OpenSIPS talking to any database, or any other component? The
idea
here being that the component is blocking and this also blocking OpenSIPS.

Yes, open source works with mysql. no problems with database detected
all incoming requests are processed quickly, also checked in real time the
occurrence of the problem

>>Also, how many child threads do you have for how many interfaces? Do you
use TCP/TLS? We have seen that sudden increases in traffic can overload the
sockets and cause backups/blocking with connection-based protocols,
especially if the remote end of a connection simply vanishes (dead peer
scenario).

No, we do not use TCP

for example, made a screenshot from the dump. in which it is clear that the
open-answer responded first to the first shoulder of the call, then created
the second one and did not react to receiving messages on it

https://drive.google.com/file/d/1X3A0Q-7njT-KLu6_7Ed1-eOnIemUULGi/view
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] opensips - stops responding to some SIP messages

[Ben Newlin]

Also, how many child threads do you have for how many interfaces? Do you use 
TCP/TLS? We have seen that sudden increases in traffic can overload the sockets 
and cause backups/blocking with connection-based protocols, especially if the 
remote end of a connection simply vanishes (dead peer scenario).

Ben Newlin

From: Users  on behalf of Alain Bieuzent 

Reply-To: OpenSIPS users mailling list 
Date: Friday, April 5, 2019 at 3:33 AM
To: OpenSIPS users mailling list 
Subject: Re: [OpenSIPS-Users] opensips - stops responding to some SIP messages

Hi Maxim,

did you take network traces at the time of the fault, does the traffic get to 
the server?
if you had memory problems that would have generated logs.

Alain

De : Users  au nom de "Maxim V. Maximov" 

Répondre à : OpenSIPS users mailling list 
Date : vendredi 5 avril 2019 à 09:26
À : 
Objet : [OpenSIPS-Users] opensips - stops responding to some SIP messages

good day

we use - Server :: OpenSIPS (2.2.6 (x86_64 / linux))
we have a powerful server. 24 gigabytes of RAM
Intel (R) Xeon (R) CPU L5520 @ 2.27GHz

we have a problem: sometimes, usually in an hour peaks or when the number of 
calls increases dramatically, by 2-3 times. the server stops responding to some 
SIP messages, at the same time some calls are successful. in 1-2 minutes 
everything is stabilized

no problems were found in the logs. no ordinary message during this period is 
received. we assumed that there might be a problem in the memory and increase 
it by 2 times.
setting the parameters STARTOPTIONS = "- u opensips -w / opensipscore -m 6024 
-M 3024" before that was m 3024 -M 1024

having restarted the server, checked and got such a problem again

we monitor our server, did not see any problems with memory or cpu on the 
server itself

Please help, how can we understand and diagnose a problem? I will be glad to 
any help

Best regards, Maxim
___ Users mailing list 
Users@lists.opensips.org 
http://lists.opensips.org/cgi-bin/mailman/listinfo/users<http://lists.opensips.org/cgi-bin/mailman/listinfo/users>
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] opensips - stops responding to some SIP messages

[David Villasmil]

Also, is OpenSIPS talking to any database, or any other component? The idea
here being that the component is blocking and this also blocking OpenSIPS.

Good luck

On Fri, 5 Apr 2019 at 11:25, Alexey Kazantsev via Users <
users@lists.opensips.org> wrote:

> Hello,
>
> please post the output of these commands:
>
>
>opensipsctl fifo log_level
>
>opensipsctl fifo get_statistics shmem:
>
>
>
> ---
> BR, Alexey
> http://alexeyka.zantsev.com/
> ___
> Users mailing list
> Users@lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
-- 
Regards,

David Villasmil
email: david.villasmil.w...@gmail.com
phone: +34669448337
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] opensips - stops responding to some SIP messages

[Alexey Kazantsev via Users]

Hello,

please post the output of these commands:


   opensipsctl fifo log_level

   opensipsctl fifo get_statistics shmem:



---
BR, Alexey
http://alexeyka.zantsev.com/
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Re: [OpenSIPS-Users] opensips - stops responding to some SIP messages

[Alain Bieuzent]

Hi Maxim,

 

did you take network traces at the time of the fault, does the traffic get to 
the server?

if you had memory problems that would have generated logs.

 

Alain

 

De : Users  au nom de "Maxim V. Maximov" 

Répondre à : OpenSIPS users mailling list 
Date : vendredi 5 avril 2019 à 09:26
À : 
Objet : [OpenSIPS-Users] opensips - stops responding to some SIP messages

 

good day

 

we use - Server :: OpenSIPS (2.2.6 (x86_64 / linux))

we have a powerful server. 24 gigabytes of RAM

Intel (R) Xeon (R) CPU L5520 @ 2.27GHz

 

we have a problem: sometimes, usually in an hour peaks or when the number of 
calls increases dramatically, by 2-3 times. the server stops responding to some 
SIP messages, at the same time some calls are successful. in 1-2 minutes 
everything is stabilized

 

no problems were found in the logs. no ordinary message during this period is 
received. we assumed that there might be a problem in the memory and increase 
it by 2 times. 

setting the parameters STARTOPTIONS = "- u opensips -w / opensipscore -m 6024 
-M 3024" before that was m 3024 -M 1024

 

having restarted the server, checked and got such a problem again

 

we monitor our server, did not see any problems with memory or cpu on the 
server itself

 

Please help, how can we understand and diagnose a problem? I will be glad to 
any help

 

Best regards, Maxim

___ Users mailing list 
Users@lists.opensips.org 
http://lists.opensips.org/cgi-bin/mailman/listinfo/users 

___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

[OpenSIPS-Users] opensips - stops responding to some SIP messages

[Maxim V. Maximov]

good day

we use - Server :: OpenSIPS (2.2.6 (x86_64 / linux))
we have a powerful server. 24 gigabytes of RAM
Intel (R) Xeon (R) CPU L5520 @ 2.27GHz

we have a problem: sometimes, usually in an hour peaks or when the number
of calls increases dramatically, by 2-3 times. the server stops responding
to some SIP messages, at the same time some calls are successful. in 1-2
minutes everything is stabilized

no problems were found in the logs. no ordinary message during this period
is received. we assumed that there might be a problem in the memory and
increase it by 2 times.
setting the parameters STARTOPTIONS = "- u opensips -w / opensipscore -m
6024 -M 3024" before that was m 3024 -M 1024

having restarted the server, checked and got such a problem again

we monitor our server, did not see any problems with memory or cpu on the
server itself

Please help, how can we understand and diagnose a problem? I will be glad
to any help

Best regards, Maxim
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users