db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Kim Haase (JIRA)" <j...@apache.org>
Subject [jira] Commented: (DERBY-3552) Handle jar files that are installed when replication is enabled
Date Mon, 31 Mar 2008 15:30:24 GMT

    [ https://issues.apache.org/jira/browse/DERBY-3552?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12583720#action_12583720
] 

Kim Haase commented on DERBY-3552:
----------------------------------

Hmmm. I think I'm still confused. Let me see if I can figure out what Jørgen's two paths
mean:

> 1: Manually install the jar files on the "used-to-be-slave" database after a failover
has occurred, or

My understanding is that failovers don't just occur; you have to initiate them. Is that correct?
So in this case, you would

 1. Install the jar(s) on the master while your app is running;
 2. Initiate a failover so that now you have two independent databases running.
 3. Install the jar(s) on the former slave.
 
However, now you have two separate databases and replication is no longer running. So if you
wanted to keep on replicating the master database, you'd have to start replication all over
again, first removing the former slave database -- that is, you'd have to do path 2.

> 2: When jar files have been installed on the master, replication is stopped and restarted
from step 1

I think this would mean you would have to

 1. Stop replication using the stopMaster attribute. This leaves both the master and slave
databases running, so you wouldn't have to do step 1 (boot the master database) in the "Starting
and running replication" topic.
 2. Stop the slave database and remove it.
 3. Freeze the master database.
 4. Copy the database (presumably with its installed jars) to the slave system.
 5. Start slave replication mode on the slave.
 6. Start master replication mode on the master.

So does this mean, basically, that you should do a failover if you don't want to replicate
the database any more and you want to have two independent databases, but you should do a
stop and restart if you want to keep replicating the master database?


> Handle jar files that are installed when replication is enabled
> ---------------------------------------------------------------
>
>                 Key: DERBY-3552
>                 URL: https://issues.apache.org/jira/browse/DERBY-3552
>             Project: Derby
>          Issue Type: Sub-task
>          Components: Replication
>    Affects Versions: 10.4.0.0, 10.5.0.0
>            Reporter: V.Narayanan
>


-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message