[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-ha
Subject:    Re: creating fail-over firewall
From:       Fabio Olive Leite <olive () conectiva ! com ! br>
Date:       2000-03-27 18:26:27
[Download RAW message or body]

Hi there,

) I wasn't aware that DRBD had reached that level of stability.  Didn't
) someone have some issues with this?
) 
) Has anyone written resource scripts so that the takeover is easy to
) implement? If so, I'd gladly include them with heartbeat.

Yes! :) There's an init script that loads the module and configures the
devices found in /etc/sysconfig/drbd/drbd{0-9} (drbd currently compiles
with support for only two devices, but that's a compile time #define), and
then waits for these devices to resync.

Then at this time (after resync) heartbeat should come up and run
/etc/ha.d/resource.d/datadisk [drbd{0-9}] {start|stop}. This script sets
the device as primary and mounts the partition (fstab should contain an
entry for the mountpoint). If only one device is specified (like
datadisk::drbd0 in haresources) then only that device is brought up,
others remain untouched. Thus it is possible to specify that two machines
will share two partitions, and the failover/failback of each shared device
can be treated independently.

If you just call datadisk start (instead of datadisk drbd{0-9} start), all
devices declared in /etc/sysconfig/drbd/ are configured as primary and
mounted. The fact that the devices can operate independently of each other
means you can have a machine serving as secondary for many others, or
whatever else.

Also, if you use Luis Claudio's patch for the "nice failback" approach, it
may not even be necessary to wait until resync finishes to start
heartbeat.

I don't think it is really necessary to include this script with
heartbeat, since they are DRBD-specific.

) What filesystem do you use on top of DRBD in your testing?

Good old ext2. :)

[]!
Fábio
( Fábio Olivé Leite   -*-   http://descartes.ucpel.tche.br/~olive )
( Linux - Distributed Systems - Fault Tolerance - Security - /etc )


------------------------------------------------------------------------------
Linux HA Web Site:
  http://linux-ha.org/
Linux HA HOWTO:
  http://metalab.unc.edu/pub/Linux/ALPHA/linux-ha/High-Availability-HOWTO.html
------------------------------------------------------------------------------

[prev in list] [next in list] [prev in thread] [next in thread]