Hi all,
Running ser0.9.7 on Solaris 10 (mysql 5.0.51 + serweb 0.9.6)
Have this problem when I stop ser service, and then start it again (ser -E -D)..
0(0) preload_udomain: ERROR: bad callid record in table aliases 0(0) preload_udomain: ERROR: for username 579981, contact sip:579981@mydomain.com 0(0) preload_udomain: ERROR: skipping...
And finally it dies with:
0(0) get_connection(): Connection not found in the pool Fallo de segmentación (core dumped)
It happens with many records on aliases table and happens too with only one record:
mysql> select * from aliases\G *************************** 1. row *************************** username: 579981 domain: 10.10.1.1 contact: sip:579981@mydomain.com received: NULL expires: 2009-02-05 11:45:21 q: 1.00 callid: The-Answer-To-The-Ultimate-Question-Of-Life-Universe-And-Everything cseq: 42 last_modified: 2009-02-05 11:45:46 replicate: 0 state: 0 flags: 128 user_agent: SIP Express Router FIFO 1 row in set (0.00 sec)
SER start normally if I delete all the records on this table. But I need create the aliases again, and this is not the normal behaviour..I think.
Here the description of aliases table:
mysql> desc aliases; +---------------+---------------------+------+-----+---------------------+-------+ | Field | Type | Null | Key | Default | Extra | +---------------+---------------------+------+-----+---------------------+-------+ | username | varchar(64) | NO | PRI | | | | domain | varchar(128) | NO | PRI | | | | contact | varchar(255) | NO | PRI | | | | received | varchar(255) | YES | | NULL | | | expires | datetime | NO | | 2020-05-28 21:32:15 | | | q | float(10,2) | NO | | 1.00 | | | callid | varchar(255) | NO | | Default-Call-ID | | | cseq | int(11) | NO | | 42 | | | last_modified | timestamp | NO | | CURRENT_TIMESTAMP | | | replicate | int(10) unsigned | NO | | 0 | | | state | tinyint(1) unsigned | NO | | 0 | | | flags | int(11) | NO | | 0 | | | user_agent | varchar(50) | NO | | | | +---------------+---------------------+------+-----+---------------------+-------+ 13 rows in set (0.01 sec)
Important: when "location" table keeps some records, ser doesn't start too:
mysql> select * from location\G *************************** 1. row *************************** username: 579981 domain: mydomain.com contact: sip:579981@10.10.60.115:5060 received: NULL expires: 2009-02-05 11:46:16 q: -1.00 callid: 1f46c06d5e034425286f095b11cf4a1a@10.10.60.115 cseq: 6 last_modified: 2009-02-05 11:45:46 replicate: 0 state: 0 flags: 0 user_agent: CM5K-TA2S (807100.02) 1 row in set (0.00 sec)
Here too I must to delete all the records to let SER start successfully.
Does anyone knows about this issue?
Regards, Claudio
caio escribió:
Hi all,
Running ser0.9.7 on Solaris 10 (mysql 5.0.51 + serweb 0.9.6)
Have this problem when I stop ser service, and then start it again (ser -E -D)..
0(0) preload_udomain: ERROR: bad callid record in table aliases 0(0) preload_udomain: ERROR: for username 579981, contact
sip:579981@mydomain.com 0(0) preload_udomain: ERROR: skipping...
[cut]
solved. there was an old installation of mysql libs (in /usr/sfw/), then I removed it, and finally recompile mysql.so module including correct mysql.h (/opt/csw/).
cheers, caio