Good point. This mysql replication stuff should probably get
a good READ me file too.
Here is what is in the vmysql.h file now:
/* Edit to match your set up
*
* If you are NOT using mysql replication:
* Then edit the UPDATE server with your mysql authentication
* information. The READ server is not used
*
* If you are using mysql replication:
* Set the UPDATE server to be your master
* and set the READ server to be your slave
*
*/
This is assuming two types of mysql setups
1) Just one server, like a standard mysql setup.
2) Mysql replication setup with a single master
and a local slave.
There isn't any code to do updates to multiple
servers. Best to leave that *major* headache
to the mysql developers.
Ken
oldarnie writes:
> Roger Arnold wrote:
>
> Hello Ken,
>
> Thanks for the reply.
> I have been reading through vpopmail-4.10.14 and am a little confused
> regarding "vmysql.h" and it's modification.
> How do you configure "vmysql.h" to use a single server instead of multiple
> servers?
> I am not that sure about mysql as a single server, so I don't want to run 2
> servers at this time.
> If I leave the configuration of the second server as it is i.e. not defined,
> and only define the first part as for a single server, will everything still
> work?
> You said that, using my config file (unchanged) everything compiled OK,
> which is what I want to do, however I am worried that without configuring
> the "vmysql.h" file for 2 servers things may go astray?
> I would apreciate some clarification, if you don't mind that is :)
>
> Thanks in advance
> Regards
> Roger
> ----- Original Message -----
> From: "Ken Jones" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> Sent: Tuesday, June 05, 2001 2:11 AM
> Subject: Re: Error trying to compile + new mysql replication release
>
>
>>
>> Try the vpopmail-4.10.14 release. I just tested it with the
>> same options you used and it compiles cleanly.
>>
>> btw: I think this version is about the cleanest we can
>> have with mysql replication.
>>
>> There is one update server defined. All sql that will
>> cause the database to be modified is sent to that server.
>>
>> There is on read only server defined, where all the
>> reads go to. If the connection to the read server fails
>> then as a last resort it will try to connect to the
>> update server.
>>
>> I think a few more code reviews are needed to make sure
>> everything is as optimized as is reasonable. And to make
>> sure none of the sql queries step on each other.
>>
>> Ken
>>
>> oldarnie writes:
>>
>> > Roger Arnold wrote:
>> >
>> > I seem to have stumbled into a major problem while trying to compile
> Vpopmail and would be obliged if someone could help ASAP please.
>> >
>> > I used the following configuration:
>> >
>> > --mandir=/usr/share/man --enable-roaming-users=y --enable-auth-logging=y
>>
>> --enable-logging=v --enable-mysql=y --enable-valias=y --enable-hardquota=3
> 000000
>> >
>> > I got the following error:
>> >
>> > vdeloldusers.c:143:2: #else without #if
>> > vdeloldusers.c:150:2: #endif without #if
>> > make[2]: *** [vdeloldusers.o] Error 1
>> > make[2]: leaving directory '/usr/src/vpopmail-4.10.9'
>> > make[1]: *** [all-recursive] Error 1
>> > make[1]: leaving directory '/usr/src/vpopmail-4.10.9'
>> > make: *** [all-recursive-am] Error 2
>> >
>> > I am compiling on a Redhat 7.1 box with an AMD K6-2 550 and 64MB Ram
>> >
>> > Any and all help would be appreciated ASAP Please
>> >
>> > Thanks in advance
>> > Regards
>> > Roger
>> >
>>
>
