Re: [HTCondor-users] forcing HA negotiator switch to stand-by central manager?

[Todd L Miller <tlmiller@xxxxxxxxxxx>]

> short question, but what is in a HA set up the best way to force a 
> negotiator daemon role to switch to the stand by central manager?

	The manual suggests this is a question based on misunderstanding 
of what HAD is trying to accomplish --

> This high availability mechanism operates by monitoring communication 
> between machines. Note that there is a significant difference in 
> communications between machines when
>
>   a machine is down
>
>   a specific daemon (the condor_had daemon in this case) is not
>   running, yet the machine is functioning
>
> The high availability mechanism distinguishes between these two, and it 
> operates based only on first (when a central manager machine is down). A 
> lack of executing daemons does not cause the protocol to choose or use a 
> new active central manager.

-- which suggests that the best way to force the active negotiator to 
switch is to:

(a) kill the active negotiator
(b) kill its replication daemon
(c) kill its HAD daemon

where step (c) is what causes the HAD daemon on the idle negotiator to 
make it active.  I think if you use `condor_off` for each step, you won't 
have problems with the master (or HAD) restarting anything:

condor_off -negotiator; condor_off -subsystem -replication; condor_off -subsystem -had

	This of course is not a permanent change.

-- ToddM