[ClusterLabs] Antw: Re: Establishing Timeouts

Christine Caulfield ccaulfie at redhat.com
Tue Oct 11 03:42:52 EDT 2016

On 11/10/16 08:22, Vladislav Bogdanov wrote:
> 11.10.2016 09:31, Ulrich Windl wrote:
>>>>> Klaus Wenninger <kwenning at redhat.com> schrieb am 10.10.2016 um
>>>>> 20:04 in
>> Nachricht <936e4d4b-df5c-246d-4552-5678653b3bd6 at redhat.com>:
>>> On 10/10/2016 06:58 PM, Eric Robinson wrote:
>>>> Thanks for the clarification. So what's the easiest way to ensure
>>>> that the
>>> cluster waits a desired timeout before deciding that a
>>> re-convergence is necessary?
>>> By raising the token (lost) timeout I would say.
>> Somewhat off-topic: I had always wished there were a kind of
>> spreadsheet where you could play with those parameters, and together
>> with required constraints you would be informed what consequences
>> changing one parameter has.
> Nice wish/idea.
> +1

It is - does anyone fancy doing it ? The code is nearly all in
totemconfig.c ;-)

totem is a complex protocol and not all of the values do what you might
think they do - if you really want to understand what is going on then
you also need to read the papers referenced at
https://github.com/corosync/corosync/wiki/Developers - particularly the
Deborah A. Agarwal one.


More information about the Users mailing list