List:General Discussion« Previous MessageNext Message »
From:Gavin Towey Date:October 7 2009 8:03pm
Subject:RE: Replication recovery
View as plain text  
B should be the only one with a bad replication position, since it was replicating when A
crashed.  So just adjust B, and A should catch up as normal (provided you have the last
24 hours of binlogs on B for A to read )

Regards,
Gavin Towey

-----Original Message-----
From: Bryan Cantwell [mailto:bcantwell@stripped]
Sent: Wednesday, October 07, 2009 11:12 AM
To: Gavin Towey
Cc: mysql@stripped
Subject: Re: Replication recovery

When only one machine dies I do send the new master position info to the
still running slave, and yes, it does the trick.
My main challenge is when A dies and is dead for 24 hours and then B
dies too. Now A is already out of synch with B and now B has a new log
position... Doesnt this make A now have a huge gap in data? How do I get
A up to date with B?

thx,
Bryancan

On 10/07/2009 12:53 PM, Gavin Towey wrote:
> In the case that one machine has a power failure, then starts a new binlog, you just
> have to set the slave to start replicating from the beginning of that binlog.  That's easy
> to detect and repair with a daemon script.  Even if both machines die, it'll be a similar
> scenario.
>
> Regards,
> Gavin Towey
>
> -----Original Message-----
> From: Bryan Cantwell [mailto:bcantwell@stripped]
> Sent: Wednesday, October 07, 2009 10:47 AM
> To: mysql@stripped
> Subject: Replication recovery
>
> I have 2 - 5.0.51a mysql databases setup in a dual master scenario. A is
> master of B and vise versa...
> In Linux 2.6.26 (if that matters).
> Everything is great while all is running normally. But, when I am
> testing the system by creating disasterous scenarios, I find some
> challenges I hope to get overcome.
>
> Let's say 'A' machine's plug gets kicked out of the wall and so when
> mysql restarts it starts fresh bin-log and the slave 'B' does not
> realize this change and we are now out of sync. 'A', however will simply
> catch up to 'B' and there MAY not be a problem.
>
> Even worse, 'A' dies and no one does anything about it, then later 'B'
> dies. Now Someone finally comes along and restarts both machines at the
> same time and neither are on the 'same page' and are totally out of sync.
>
> How, without re-copying the datafiles and starting over (after
> determining the most up to date machine to use), can I bring both 'A'
> and 'B' to the same point so I can move forward?
>
> Thanks,
>
> Bryancan
>
> --
> MySQL General Mailing List
> For list archives: http://lists.mysql.com/mysql
> To unsubscribe:    http://lists.mysql.com/mysql?unsub=1
>
>
> The information contained in this transmission may contain privileged and
> confidential information. It is intended only for the use of the person(s) named above. If
> you are not the intended recipient, you are hereby notified that any review,
> dissemination, distribution or duplication of this communication is strictly prohibited.
> If you are not the intended recipient, please contact the sender by reply email and
> destroy all copies of the original message.
>


The information contained in this transmission may contain privileged and confidential
information. It is intended only for the use of the person(s) named above. If you are not
the intended recipient, you are hereby notified that any review, dissemination,
distribution or duplication of this communication is strictly prohibited. If you are not
the intended recipient, please contact the sender by reply email and destroy all copies
of the original message.
Thread
Replication recoveryBryan Cantwell7 Oct
  • RE: Replication recoveryGavin Towey7 Oct
    • Re: Replication recoveryBryan Cantwell7 Oct
      • RE: Replication recoveryGavin Towey7 Oct