[Linux-ha-dev] checkpointd and multiple nodes
alanr at unix.sh
Sat Oct 16 10:21:22 MDT 2004
Alan Robertson wrote:
> Kristen Smith wrote:
>> Is the current checkpointd implementation (1.2.3) meant strictly for 2
>> clusters? I was hoping to use it on a 2+ node cluster to keep data in
>> but after playing around with it a bit, I believe it might be currently
>> written for 2 nodes only. Am I just confused or is my assumption correct?
> If you use heartbeat's resource management code (that is, have a
> non-empty haresources file), then it *heartbeat's resource management*
> will disappoint you - probably quite badly.
> I also think the Intel folks have been working more on the HEAD
> (unstable) branch than on 1.2 (stable) branch. I'm pretty sure they
> have a lot of code that they didn't port back to 1.2.
> IIRC, if you get the right version of the code, and use it without
> enabling heartbeat's resource management, then you should be OK.
Sorry, I forgot to say "on larger clusters" as the qualifier for everything
I said above.
Alan Robertson <alanr at unix.sh>
"Openness is the foundation and preservative of friendship... Let me claim
from you at all times your undisguised opinions." - William Wilberforce
More information about the Linux-HA-Dev