I have noticed a consistent difference between the location records OpenSER has cached and those stored in the database.
e.g. openserctl ul show --brief database engine 'MYSQL' loaded Control engine 'FIFO' loaded entering fifo_cmd ul_dump brief Domain:: location table=512 records=19 max_slot=2
Compared to: select count(*) from ser.location; +----------+ | count(*) | +----------+ | 10 | +----------+ 1 row in set (0.00 sec)
This difference is consistent over time.
When I look at the database logs OpenSER is trying to update an existing record in the database which, for some reason, doesn't exist.
Our config has the following relevant parameters modparam("usrloc", "db_mode", 2)
Openser 1.3.2
Any thoughts on how to prevent this circumstance from occurring and how to synchronise the cache with the db in an orderly manner? I had thought that after 120 seconds (timer_interval default)this would happen but it doesn't seem to.
Any advice appreciated.
Regards
Cameron
Hello,
On 10/23/08 02:28, CSB wrote:
I have noticed a consistent difference between the location records OpenSER has cached and those stored in the database.
e.g. openserctl ul show --brief database engine 'MYSQL' loaded Control engine 'FIFO' loaded entering fifo_cmd ul_dump brief Domain:: location table=512 records=19 max_slot=2
Compared to: select count(*) from ser.location; +----------+ | count(*) | +----------+ | 10 | +----------+ 1 row in set (0.00 sec)
This difference is consistent over time.
When I look at the database logs OpenSER is trying to update an existing record in the database which, for some reason, doesn't exist.
Our config has the following relevant parameters modparam("usrloc", "db_mode", 2)
Openser 1.3.2
Any thoughts on how to prevent this circumstance from occurring and how to synchronise the cache with the db in an orderly manner? I had thought that after 120 seconds (timer_interval default)this would happen but it doesn't seem to.
indeed seems a problem. Perhaps the insert of the record failed, therefore the update fails as well. Reason can be that the DB was not running when openser tried to insert. Perhaps in this case is indicated an attempt to insert.
This should be registered on tracker too, not to forget about, could you please do it:
http://sourceforge.net/tracker/?group_id=139143
Cheers, Daniel