Hi, what you mean for 2-way replication in mysql?? The replica as A-->B-->A or the clustering of mysql 5?? I see limitation in both situation:
In A-->B-->A you have trouble in case of failure of one server to get again a working environment being sure that all is syncrhonized well. If you can suggest a good way to accomplish it will be welcome. I'm trying to do the same thing.
As i know (may be i'm wrong and i will be happy to be) In clustering you do not have support for fereign keys and there are other kind of limitation.
Let me know if you can suggest a good way to do it.
I need the transactional functions, foreign keys and to be sure that data are flushed to disk almost immediately after the query executed (this is something that the cluster do not assure to you because it keep in memory lot of data and flush it to disk in big chunks) for the application that run on top of SER/OpenSer.
The goal of my application is to have 2 OpenSer (balanced using SRV records or something like this) servers working on two separate mysql but dealing with the same data. For now the best solution for reliability that i found is to have 2 OpenSer servers to operate on the 1st mysql and the 2nd in sync as slave with one way replication and in case of failure to move the IP address of mysql on the slave and restart the slave as master. All this with Heartbeat package.
Bye, Marcello
Hi Daryl,
the 2-way replication in Mysql works and it might be used along with OpenSER - depends of your configuration.
regards, bogdan
Daryl Sanders wrote:
Hi Marcello,
you may try a mix replication: 2 way DB replication between mysql servers. 2 way SIP replication between IP servers using t_replicate()
regards, bogdan
Marcello Lupo wrote:
Hi, what you mean for 2-way replication in mysql?? The replica as A-->B-->A or the clustering of mysql 5?? I see limitation in both situation:
In A-->B-->A you have trouble in case of failure of one server to get again a working environment being sure that all is syncrhonized well. If you can suggest a good way to accomplish it will be welcome. I'm trying to do the same thing.
As i know (may be i'm wrong and i will be happy to be) In clustering you do not have support for fereign keys and there are other kind of limitation.
Let me know if you can suggest a good way to do it.
I need the transactional functions, foreign keys and to be sure that data are flushed to disk almost immediately after the query executed (this is something that the cluster do not assure to you because it keep in memory lot of data and flush it to disk in big chunks) for the application that run on top of SER/OpenSer.
The goal of my application is to have 2 OpenSer (balanced using SRV records or something like this) servers working on two separate mysql but dealing with the same data. For now the best solution for reliability that i found is to have 2 OpenSer servers to operate on the 1st mysql and the 2nd in sync as slave with one way replication and in case of failure to move the IP address of mysql on the slave and restart the slave as master. All this with Heartbeat package.
Bye, Marcello
Hi Daryl,
the 2-way replication in Mysql works and it might be used along with OpenSER - depends of your configuration.
regards, bogdan
Daryl Sanders wrote:
Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users
-
Bogdan-Andrei Iancu -
Marcello Lupo