Re: [OpenSIPS-Devel] [opensips] Core dump at db_mysql (#186)

[apsaras]

(gdb) frame 0
#0  0x2b0f5e34fd99 in db_mysql_get_columns (_h=, 
_r=0x2b0f5e14bb70) at res.c:71
71  RES_NAMES(_r)[col]->len = strlen(fields[col].name);
(gdb) print fields[1]
Cannot access memory at address 0x78
(gdb) print *_r
$1 = {col = {names = 0x2b0f5e1769f0, types = 0x2b0f5e176c48, n = 25}, rows = 
0x0, n = 0, res_rows = 0, last_row = 0}


---
Reply to this email directly or view it on GitHub:
https://github.com/OpenSIPS/opensips/issues/186#issuecomment-38857636___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] Core dump at db_mysql (#186)

[Liviu Chircu]

Go to frame zero with:

frame 0

Print useful stuff with:

print fields[1]
print *_r

---
Reply to this email directly or view it on GitHub:
https://github.com/OpenSIPS/opensips/issues/186#issuecomment-38857028___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] drouting: failed to work with nonSQL database (#173)

[Ovidiu Sas]

One solution would be to change the id container inside the module from int
to str.
For most common databases, the id will be an int that will be converted
into a str and stored as a str.
For mongo, the id is retrieved as an ObjectId and it is converted and
presented to the upper module as str, so storing it as a str would be
'native'.

This requires changes in all the modules that are explicitly using the
primay key 'id'.
On the other hand, there are only a few modules that require explicit
access to 'id'.

For now, this is the only solution that I can think of.

-ovidiu


On Thu, Mar 27, 2014 at 2:21 PM, Bogdan Andrei IANCU <
notificati...@github.com> wrote:

> @ovidiusas  - unfortunately not only the
> drouting module is affected by this problem - there are other modules using
> the "id" primary key (even storing it) - so I would rather look for a
> global solutions on this. Do you have any suggestions ?
>
> --
>
> Reply to this email directly or view it on 
> GitHub
> .
>
>
> ___
> Devel mailing list
> Devel@lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/devel
>
>


-- 
VoIP Embedded, Inc.
http://www.voipembedded.com
___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] Core dump at db_mysql (#186)

[apsaras]

Hello Liviuchircu
No startup options
Fresh setup, no data in the database
That is what I get at log
Mar 27 21:53:07 sip kernel: opensips[727]: segfault at  rip 
2ad6859ced99 rsp 74511cd0 error 4

Hello Bogdan
How can I do that?

---
Reply to this email directly or view it on GitHub:
https://github.com/OpenSIPS/opensips/issues/186#issuecomment-38852512___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] changing db_text database causes coredump (#180)

[Bogdan Andrei IANCU]

It seems like, but I was trying to understand how "OpenSIPS" decided to 
"re-load" the file

---
Reply to this email directly or view it on GitHub:
https://github.com/OpenSIPS/opensips/issues/180#issuecomment-38846046___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] drouting: failed to work with nonSQL database (#173)

[Bogdan Andrei IANCU]

@ovidiusas - unfortunately not only the drouting module is affected by this 
problem - there are other modules using the "id" primary key (even storing it) 
- so I would rather look for a global solutions on this. Do you have any 
suggestions ?

---
Reply to this email directly or view it on GitHub:
https://github.com/OpenSIPS/opensips/issues/173#issuecomment-38841578___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] changing db_text database causes coredump (#180)

[Ovidiu Sas]

It seems that this was related to a race between the text file being
updated by an external script and opensips loading the incomplete file.
On Mar 27, 2014 2:13 PM, "Bogdan Andrei IANCU" 
wrote:

> @lbarth  , what db_mode are you using for the
> db_text module ? 1 ?
>
> --
> Reply to this email directly or view it on 
> GitHub
> .
>
> ___
> Devel mailing list
> Devel@lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/devel
>
>
___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] changing db_text database causes coredump (#180)

[Bogdan Andrei IANCU]

@lbarth , what db_mode are you using for the db_text module ? 1 ?

---
Reply to this email directly or view it on GitHub:
https://github.com/OpenSIPS/opensips/issues/180#issuecomment-38840697___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] Core dump at db_mysql (#186)

[Bogdan Andrei IANCU]

@apsaras, could you please, in frame 0, print fields[1] and *_r also ?

Thanks and regards,
Bogdan

---
Reply to this email directly or view it on GitHub:
https://github.com/OpenSIPS/opensips/issues/186#issuecomment-38840083___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [opensips] Core dump at db_mysql (#186)

[Liviu Chircu]

Do you start OpenSIPS with the default packaged memory size? (i.e. no "-M" 
parameter)

If yes, did you happen to have a full dialog table when you started? Do you see 
any errors in the OpenSIPS logs?

The latest memory fixes of today did not cover the standard memory allocators, 
as they behaved very well with my tests, at least. So we need to look at the 
problem from all sides.

Liviu

---
Reply to this email directly or view it on GitHub:
https://github.com/OpenSIPS/opensips/issues/186#issuecomment-38839440___
Devel mailing list
Devel@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/devel

Re: [OpenSIPS-Devel] [OpenSIPS-Users] [RELEASE] OpenSIPS 1.11 beta major release is out

[Bogdan-Andrei Iancu]

Liviu will prepare some paper soon on these performance improvement from 
1.10 to 1.11. This new memory manager is distributed and is able to 
scale with the number of processes/core you have ; the existing 
implementations proved to be a bottleneck (limiting the degree of 
parallelism you could achive) because of the monolithic locking. With 
this new mem manager we were able to scale up to 50K cps with a single 
opensips instance.


Regards,

Bogdan-Andrei Iancu
OpenSIPS Founder and Developer
http://www.opensips-solutions.com

On 24.03.2014 18:46, Ryan Bullock wrote:
This is an exciting release! The new memory allocator looks really 
interesting. Are there any numbers showing how its performance 
compares to the current?


Great work!


~Ryan Bullock


On Thu, Mar 20, 2014 at 1:29 PM, Bogdan-Andrei Iancu 
mailto:bog...@opensips.org>> wrote:


Hello everyone!

The OpenSIPS Project is proud to announce the release of OpenSIPS
version 1.11 (beta)!

We would like to thank the OpenSIPS community for all of their
hard and diligent work in making this release possible. We could
never have done it without you!

Special thanks go out to Ovidiu Sas, Walter Doekes, Damien Sandrs,
Nick Altmann, Brett Nemeroff, Ryan Bullock (and many others) for
your amazing contribution on this release. We truly appreciate you!

Building on our industry ready platform, we're excited to
introduce many new features and updates. But also we've continued
to make inroads in developing an easier to use OpenSIPS.

Version 1.11 brings with it enhancements to the core, script
handling, and many important modules.

The OpenSIPS core has received a new memory allocator to increase
performance. It's tunable and provides fine-grained locking!

We've also heard your requests on improving scripting
capabilities. Say hello to the SCRIPT_HELPER module and to the
"for-each" statement! The learning curve will not be so steep again!

We've also introduced 4 new modules in this release...

- B2B_SCA module providing new shared call appearance features
- CALL_CENTER module that introduces call queue features
- MI_JSON module to encode data in JSON format over HTTP for the
MI Interface commands
- SCRIPT_HELPER module to simplify the script/configuration for
beginners

In all, too many features to list. However you can view them all
by visiting the version page at:
http://www.opensips.org/About/Version-1-11-0

Again, we are excited about all the new changes version 1.11
delivers. We continue to appreciate all the feedback and help from
the community.

We still have many things to be done to get to the stable release
(in ~1 month), like improving the documentation, keep working on
fixing bugs, excessive testing and others.

We're always listening to your requests, so never be shy in making
one!


Many thanks,
The OpenSIPS Project Team

-- 
Bogdan-Andrei Iancu

OpenSIPS Founder and Developer
http://www.opensips-solutions.com


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




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


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

[OpenSIPS-Devel] [OpenSIPS/opensips] edd872: call_center: fix some db schema errors

[Liviu Chircu]

  Branch: refs/heads/master
  Home:   https://github.com/OpenSIPS/opensips
  Commit: edd872182882536f3cf37bd5c4be633075f265d4
  
https://github.com/OpenSIPS/opensips/commit/edd872182882536f3cf37bd5c4be633075f265d4
  Author: Liviu Chircu 
  Date:   2014-03-27 (Thu, 27 Mar 2014)

  Changed paths:
M db/schema/cc_agents.xml
M db/schema/cc_cdrs.xml
M db/schema/cc_flows.xml
M scripts/db_berkeley/opensips/cc_agents
M scripts/db_berkeley/opensips/cc_cdrs
M scripts/db_berkeley/opensips/cc_flows
M scripts/dbtext/opensips/cc_cdrs
M scripts/mysql/call_center-create.sql
M scripts/oracle/call_center-create.sql
M scripts/pi_http/call_center-mod
M scripts/pi_http/call_center-table
M scripts/pi_http/pi_framework.xml
M scripts/postgres/call_center-create.sql

  Log Message:
  ---
  call_center: fix some db schema errors

* reduce the cc_agents skills column size
* fix some primary key issues in cc_flows and cc_agents
* fix a couple of bad column names in cc_cdrs


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

[OpenSIPS-Devel] [OpenSIPS/opensips] 51c151: call_center: fix some db schema errors

[Liviu Chircu]

  Branch: refs/heads/1.11
  Home:   https://github.com/OpenSIPS/opensips
  Commit: 51c1518be82cea753caa72d49e57f7a641e2eab1
  
https://github.com/OpenSIPS/opensips/commit/51c1518be82cea753caa72d49e57f7a641e2eab1
  Author: Liviu Chircu 
  Date:   2014-03-27 (Thu, 27 Mar 2014)

  Changed paths:
M db/schema/cc_agents.xml
M db/schema/cc_cdrs.xml
M db/schema/cc_flows.xml
M scripts/db_berkeley/opensips/cc_agents
M scripts/db_berkeley/opensips/cc_cdrs
M scripts/db_berkeley/opensips/cc_flows
M scripts/dbtext/opensips/cc_cdrs
M scripts/mysql/call_center-create.sql
M scripts/oracle/call_center-create.sql
M scripts/pi_http/call_center-mod
M scripts/pi_http/call_center-table
M scripts/pi_http/pi_framework.xml
M scripts/postgres/call_center-create.sql

  Log Message:
  ---
  call_center: fix some db schema errors

* reduce the cc_agents skills column size
* fix some primary key issues in cc_flows and cc_agents
* fix a couple of bad column names in cc_cdrs
(cherry picked from commit edd872182882536f3cf37bd5c4be633075f265d4)


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

[OpenSIPS-Devel] [OpenSIPS/opensips] 8b8ab7: hp malloc: reorder some initializing operations

[Liviu Chircu]

  Branch: refs/heads/1.11
  Home:   https://github.com/OpenSIPS/opensips
  Commit: 8b8ab79a19d90c173a682ff373003ec2d32b41c3
  
https://github.com/OpenSIPS/opensips/commit/8b8ab79a19d90c173a682ff373003ec2d32b41c3
  Author: Liviu Chircu 
  Date:   2014-03-27 (Thu, 27 Mar 2014)

  Changed paths:
M mem/hp_malloc.c
M mem/shm_mem.c

  Log Message:
  ---
  hp malloc: reorder some initializing operations

* hp_mem_warming() fails if anything is allocated before
(it cannot access the big fragment anymore)
(cherry picked from commit 9f710642b9da0d091d67b79b95b4a89cac954315)


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

[OpenSIPS-Devel] [OpenSIPS/opensips] 9f7106: hp malloc: reorder some initializing operations

[Liviu Chircu]

  Branch: refs/heads/master
  Home:   https://github.com/OpenSIPS/opensips
  Commit: 9f710642b9da0d091d67b79b95b4a89cac954315
  
https://github.com/OpenSIPS/opensips/commit/9f710642b9da0d091d67b79b95b4a89cac954315
  Author: Liviu Chircu 
  Date:   2014-03-27 (Thu, 27 Mar 2014)

  Changed paths:
M mem/hp_malloc.c
M mem/shm_mem.c

  Log Message:
  ---
  hp malloc: reorder some initializing operations

* hp_mem_warming() fails if anything is allocated before
(it cannot access the big fragment anymore)


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

[OpenSIPS-Devel] [OpenSIPS/opensips] 5663e4: hp malloc: fix shm statistics

[Liviu Chircu]

  Branch: refs/heads/1.11
  Home:   https://github.com/OpenSIPS/opensips
  Commit: 5663e4129bb2a6aced49715a57d50406e0968a6f
  
https://github.com/OpenSIPS/opensips/commit/5663e4129bb2a6aced49715a57d50406e0968a6f
  Author: Liviu Chircu 
  Date:   2014-03-27 (Thu, 27 Mar 2014)

  Changed paths:
M mem/hp_malloc.c
M mem/hp_malloc.h
A mem/hp_malloc_stats.c
A mem/hp_malloc_stats.h
M mem/shm_mem.c

  Log Message:
  ---
  hp malloc: fix shm statistics

* processes now update fragment statistics for each bucket
* the MI workers iterate and aggregate all memory information
* the built-in maximum sampling rate is 200ms
- repeating a fetch in < 200ms will return the same results
* this patch also corrects a rare deadlock in hp_realloc
- hp_realloc "growth" may deadlock with hp_malloc "shrink"
- the best solution is to drop the "growth" realloc operation
- this also merges well with the allocator fragmentation 
requirements
(cherry picked from commit 9ad2947eed7add4510b1fa388004524cc50d5bab)


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

[OpenSIPS-Devel] [OpenSIPS/opensips] 9ad294: hp malloc: fix shm statistics

[Liviu Chircu]

  Branch: refs/heads/master
  Home:   https://github.com/OpenSIPS/opensips
  Commit: 9ad2947eed7add4510b1fa388004524cc50d5bab
  
https://github.com/OpenSIPS/opensips/commit/9ad2947eed7add4510b1fa388004524cc50d5bab
  Author: Liviu Chircu 
  Date:   2014-03-27 (Thu, 27 Mar 2014)

  Changed paths:
M mem/hp_malloc.c
M mem/hp_malloc.h
A mem/hp_malloc_stats.c
A mem/hp_malloc_stats.h
M mem/shm_mem.c

  Log Message:
  ---
  hp malloc: fix shm statistics

* processes now update fragment statistics for each bucket
* the MI workers iterate and aggregate all memory information
* the built-in maximum sampling rate is 200ms
- repeating a fetch in < 200ms will return the same results
* this patch also corrects a rare deadlock in hp_realloc
- hp_realloc "growth" may deadlock with hp_malloc "shrink"
- the best solution is to drop the "growth" realloc operation
- this also merges well with the allocator fragmentation 
requirements


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