Hi!

I have thought about following "solutions":
1) kernel software watchdog and userland retriggering
2) simple keepalive daemon, so slave can take over(and eventually fail too) if 
master fails

Your proposal would also be somewhat good results same failure as my 2nd 
proposal.


Rivo

On 26/04/2017, 00:48, "Stuart Henderson" <[email protected]> wrote:

    On 2017/04/25 21:31, Rivo Nurges wrote:
    > Unfortunately its still able to send carp hellos and remains carp master.
    
    Sorry not an answer to your problem, but wouldn't it be nice to have a
    "carp watchdog" that must be retriggered by a userland program...
    
    

Reply via email to