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

List:       linux-ha-dev
Subject:    Re: [Linux-ha-dev] node deletion
From:       Guochun Shi <gshi () ncsa ! uiuc ! edu>
Date:       2005-10-25 15:18:03
Message-ID: 435E4CAB.8010807 () ncsa ! uiuc ! edu
[Download RAW message or body]

Alan Robertson wrote:

> Guochun Shi wrote:
>
>>> >>What would simplify it is if you could say that you can never 
>>> delete a node unless all nodes except the one(s) being deleted are 
>>> up. Having the deleted nodes up in heartbeat doesn't particularly 
>>> hurt anything - in fact it helps.
>>> >>
>>> >>That assumption _would_ simplify things.
>>> >
>>> >
>>> > That would mean we cannot delete a node, if two nodes crashed and
>>> > cannot be recovered. I don't think it is acceptable in practice.
>>>
>>> That was my first thought. But I decided I was wrong :-)
>>>
>>> So, here's the scenario... You have a 6-node cluster, and two machines
>>> died due to a lightning strike (or something).
>>>
>>> You now issue a _single_ command to delete both nodes. Since all the
>>> nodes except these two are up, then the command should be accepted.
>>>
>>> Does that work?
>>
>>
>>
>> That will work. But you might want to only delete one node even two 
>> nodes crashes.
>> (You can delete both nodes, then add the other node though)
>
>
> Or you can bring up the other node, and then delete the one node.
>
> But, it would simplify things a lot.
>
> What do people think:
>     Is this restriction acceptable?
>
I am happy to see the restriction :)

-Guochun
_______________________________________________________
Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/
[prev in list] [next in list] [prev in thread] [next in thread]